Skip to main content

Full text of "The University of Kansas science bulletin"

See other formats


v ■ o\ *n 



HARVARD UNIVERSITY 




LIBRARY 



OF THE 



Museum of Comparative Zoology 



The Library -' 
Museum of Comparative Zoology 
Harvard University 



THE UNIVERSITY OF KANSAS 

SCIENCE BULLETIN 

MUS. CO MP. ZOOS 
LIBRARY 

FEB 25 1972 



HARVARD 
UNIVERSITY 



REVISION OF THE BEE GENUS AGAPOSTEMON 
(HYMENOPTERA: HALICTIDAE) 

By 

Radclyffe B. Roberts 



Vol. XLIX Pages 437-590 Feb. 16, 1972 No. 9 



S-N4- Lcatuv 



ANNOUNCEMENT 

The University of Kansas Science Bulletin (continuation of the Kansas Uni- 
versity Quarterly) is an outlet for scholarly scientific investigations carried out 
at the University of Kansas or by University faculty and students. Since its 
inception, volumes of the Bulletin have been variously issued as single bound 
volumes, as two or three multi-paper parts or as series of individual papers. 
In each case, issuance is at irregular intervals, with each volume approximately 
1000 pages in length. 

The supply of all volumes of the Kansas University Quarterly is now ex- 
hausted. However, most volumes of the University of Kansas Science Bulletin 
are still available and are offered, in exchange for similar publications, to learned 
societies, colleges and universities and other institutions, or may be purchased 
at $15.00 per volume. Where some of these volumes were issued in parts, in- 
dividual parts are priced at the rate of 1% cents per page. Current policy, ini- 
tiated with volume 46, is to issue individual papers as published. Such separata 
may likewise be purchased individually at the rate of 1H cents per page. Sub- 
scriptions for forthcoming volumes may be entered at the rate of $15.00 per 
volume. All communications regarding exchanges, sales and subscriptions should 
be addressed to the Exchange Librarian, University of Kansas Libraries, 
Lawrence, Kansas 66044. 

Reprints of individual papers for personal use by investigators are available 
gratis for most recent and many older issues of the Bulletin. Such requests should 
be directed to the author. 



Editor 
Charles R. Wyttenbach 



Editorial Board 

Kenneth B. Armitage 
Richard F. Johnston 

Paul A. Kitos 

Charles D. Michener 

Delbert M. Shankel 

George W. Byers, Chairman 



THE UNIVERSITY OF KANSAS 

SCIENCE BULLETIN 

Vol. XLIX Pages 437-590 Feb. 16, 1972 No. 9 



TABLE OF CONTENTS 

ABSTRACT 439 

INTRODUCTION 439 

ACKNOWLEDGMENTS 440 

HISTORICAL REVIEW 441 

EXCLUDED SPECIES 442 

GENERIC DIAGNOSIS 444 

INFRAGENERIC STRUCTURE 444 

ZOOGEOGRAPHY 446 

METHODS 449 

KEY TO SPECIES 

United States and Canada 451 

Meso America 453 

West Indies 456 

South America 458 

DESCRIPTIONS, SYNONYMIES AND DISTRIBUTIONS 

aenigma 460 

alayoi 461 

angelicus 461 

ascius 463 

atrocaerulens 464 

boliviensis 467 

centratus 469 

chapadensis 470 

coc\erelli 471 

coloradinus 473 

columbi 476 

cubensis 478 

cyaneus 479 

erebus 480 

femoratus 482 

heterurus 483 

hispaniolicus 484 

inca 484 

insularis 485 



intermedins 486 

jamaicensis 490 

kphliellus 491 

lanosus 494 

leunculus 495 

melliventris 499 

mexicanus 503 

mourei 504 

nasutus 508 

obscuratus 513 

ochromops 513 

peninsularis 515 

poeyi 517 

radiatus 520 

rhopalocera 523 

sapphirinus 524 

semimelleus 525 

splendens 529 

swainsonae 531 

texanus 533 

tyleri 542 

viequesensis 544 

virescens 547 

viridulus 550 

NOMINA DUBIA 554 

APPENDIX A 555 

LITERATURE CITED 556 

FIGURES 559 

INDEX TO INCLUDED TAXA 588 



Revision of the Bee Genus Agapostemon 
(Hymenoptera: Halictidae) 



1.2 



Radclyffe B. Roberts 3 



ABSTRACT 



In this study 43 species of Agapostemon are recognized from the Americas. 
Of these, the following 19 species are new: A. aenigma, A. alayoi, A. ascius, 
A. boliviensis, A. columbi, A. cubensis, A. cyaneus, A. erebus, A. hispaniolicus, 
A. inca, A. insularis, A. intermedins, A. jamaicensis, A. lanosus, A. mexicanus, 
A. mourei, A. ochromops, A. peninsnlaris, A. sapphirmus. Forty-six names are 
listed as synonyms, 26 for the first time. Twenty-two species previously placed 
in Agapostemon but now placed in other genera are listed. 

Separate keys are provided for species from America north of Mexico, Meso- 
America, The West Indies and South America. All species are described and 
most are illustrated. The geographic distribution and variation of the species 
and species groups are discussed, with particular regard to speciation of North 
American groups. 

INTRODUCTION 

The genus Agapostemon occurs only in the Western Hemisphere where 
it ranges from southern Canada to Paraguay. It is the only member of a 
group of allied genera to be found north of Mexico — South America being 
the center of abundance of most of its relatives. 

Agapostemon is polythetic and cannot be differentiated from related 
genera on the basis of any one character or group of characters. Neverthe- 
less, most species are characterized by their metallic green or blue head and 
mesosoma and their contrasting black or black and yellow metasoma. 

Until recently, the biology of this genus was poorly known. However, in 
1969 Eickwort and Eickwort described in detail the nesting and foraging 
behavior of A. nastitus in Central America. In the same year I reported on 
the biology of the North American A. radiatus, A. splendens and A. texanus 
and summarized available information on other species. These biological 
investigations have complemented the present classification. 

The primary objective of this revision is to describe and classify the 
species of Agapostemon and to elucidate their evolutionary history. It is 
possible to recognize species groups and in some cases provide reasonable 

1 Contribution number 1452 from the Department of Entomology, University of Kansas, 
Lawrence, Kansas. 

"This study was supported by National Science Foundation Grant GB 91 to the University 
of Kansas (C. D. Michener, principal investigator). 

3 Present address: Department of Entomology, Oregon State University, Corvallis, Oregon 
97330. 



440 The University Science Bulletin 

explanations of their origins, but the number of species with no apparent 
affinities proscribes extensive speculation on the phylogeny of the genus. 

A second objective of this work is to facilitate identification of species 
of Agapostemon. To this end keys, descriptions and illustrations have been 
prepared in as simple and uniform a style as possible. Species descriptions 
are in alphabetical sequence, because I believe a phenetic or "phylogenetic" 
sequence presumes too much knowledge on the part of the readers. The 
keys are regional and artificial for the sake of brevity and clarity respectively. 

ACKNOWLEDGMENTS 

I would like to express my gratitude to Dr. C. D. Michener for pointing out 
the need for this study; for examining specimens in Washington, London, Genoa, 
and in Pretoria, South Africa; and for his interest, encouragement, and guidance 
during the course of this study. Special thanks are also extended to Dr. G. W. 
Byers for his very helpful advice and careful editing of the manuscript. 

The taxonomic portion of this study was made possible by the efforts of the 
curators, graduate students and others who selected and loaned nearly 50,000 
specimens from the following collections: Academy of Natural Sciences of 
Philadelphia; American Museum of Natural History; Brigham Young Univer- 
sity; British Museum (Natural History); California Department of Agriculture 
(Sacramento); Canadian National Collection; Carnegie Museum; Chicago 
Natural History Museum; Colorado State University; Cornell University; Florida 
State Collection of Arthropods, Gainesville; Illinois Natural History Survey; 
Iowa State University; Instituto e Museo di Zoologia Universita di Torino; 
Kansas State University; Los Angeles County Museum; Michigan State Univer- 
sity; Milwaukee Public Museum; Montana State College; Museo Civico di Storia 
Naturale (Genoa); Museum National d'Histoire Naturelle (Paris); Museum of 
Comparative Zoology; Naturhistorisches Museum (Vienna); North Carolina 
State of the University of North Carolina; Northwestern State College (Louisi- 
ana); Ohio State University; Oklahoma State University; Oregon State Uni- 
versity; Pennsylvania State University; Purdue University; Riksmuseum van 
Natuurlijke Historie (Amsterdam); Rutgers, The State University; San Jose 
State College; South Dakota State College; Stanford University; Transvaal Mu- 
seum (Pretoria, South Africa); United States Department of Agriculture, Wild 
Bee Pollination Investigations (Logan, Utah); United States National Museum; 
University of Arizona; University of Arkansas; University of California at 
Berkeley, Davis, and Riverside; University of Colorado; University of Georgia; 
University of Idaho; University of Kansas; University of Louisville (Kentucky); 
University of Michigan; University of Minnesota; University of Missouri; Uni- 
versity of Nebraska; University of Nevada; University of North Dakota; Univer- 
sity of Tennessee; University of Wisconsin; University of Wyoming; Utah State 
University; Washington State University; Zoologisches Museum der Humboldt- 
Universitat (Berlin). 

The following individuals graciously loaned specimens from their personal 
collections: Pastor Alayo D., Havana, Cuba; the late R. R. Dreisbach (collection 



Revision of the Bee Genus Agapostemon 441 

now at Michigan State University); R. A. Morse, Cornell University; D. W. 
Ribble, University of Wyoming; and G. I. Stage, United States National Museum. 
Through their cooperation and hospitality the following individuals did much 
to make my visits to museums both profitable and pleasant: P. H. Arnaud, Jr., 
California Academy of Sciences; G. E. Bohart, U.S.D.A., Logan, Utah; K. V. 
Krombein and G. I. Stage, U.S. National Museum; J. G. Rozen, Jr., American 
Museum of Natural History; the late J. A. G. Rehn and the late H. J. Grant, 
Academy of Natural Sciences of Philadelphia. 

Finally, I am grateful to my wife, Guinnevere, whose multifarious talents as 
collector, observer, secretary, editor and counselor have contributed immeasurably 
to the completion of this study. 

HISTORICAL REVIEW 

The taxon Agapostemon was first proposed in 1844 by F. E. Guerin- 
Meneville as a subgenus of Andrena in the following statements: 

Nous connaissons plusieurs especies a cuisses ainsi renflees. Ce 
sont des males. Peut-etre jugera-t-on a propos de les reunir en un 
sous-genre, que nous proposerions de nommer Agapostemon. II 
serait aux Andrenes ce qu'est le genre Nomia parmi les Halictes. 

The only included species was Andrena (Agapostemon) femoralis 
Guerin 1844. Dalla Torre (1896) correctly recognized this species to be a 
junior synonym of Apis viridula Fabricius 1793. Therefore the correct name 
for the type species of the genus Agapostemon is now Agapostemon viridulus 
(Fabricius). Agapostemon was first described and accorded generic rank 
in 1853 by Frederick Smith, who listed seven species, four of them new. 
In 1896 Dalla Torre catalogued 16 species and placed four species in 
synonymy. He also gave the Latin translation of Guerin's transliterated 
Greek compound, Agapostemon, as "aya7ra<o amo, aT-q/jLon* stemen" In 
English the translation is "lover of stamens." 

In 1897 Robertson redescribed the species known from the United States 
and listed their synonyms. The first key, published by Titus in 1901, included 
only the species known to occur in Colorado, and in November of the same 
year Crawford revised the North American species of Agapostemon. In 
this revision Crawford re-described the genus, described seven new species, 
and included a key to the 15 species then known from North America. In 
1902 Robertson published keys to the genera and species of the North 
American Halictinae in which he commented on the relationships of the 
genera. Although he cited very few characters, the keys and generic con- 
cepts were remarkably good. 

In 1903 Vachal reduced Agapostemon to subgeneric rank in Halictus. 
He included 28 species (nine of them new) in the subgenus Agapostemon 
and proposed the new subgenus Paragapostemon for 25 additional species 



442 The University Science Bulletin 

with hairy eyes and without a complete propodeal carina. Although his 
keys were good, Vachal's conservative generic concepts were not widely 
accepted, and Cockerell (1905) accorded generic rank to both Agapostemon 
and Paragapostemon. Vachal did not cite a type species for Paragapostemon, 
but Cockerell (1905, in footnote) designated Halictus (Paragapostemon) 
podager Vachal as the type species. 

Schrottky (1909a) erected the genus Pseudagapostemon (type species 
Agapostemon arenarius Schrottky), thereby removing most of the species 
from Paragapostemon as well as additional species from Agapostemon. 
Schrottky published another paper (1909b) in which he reduced seven 
species of South American Agapostemon to synonymy and transferred four 
more from Agapostemon to Pseudagapostemon. In 1918(a) Cockerell re- 
duced Pseudagapostemon to subgeneric rank in Agapostemon, but never re- 
ferred to it as such thereafter. 

In 1936 Sandhouse revised the species of Agapostemon occurring in the 
United States. Although she examined about 4,000 specimens, her work 
does not make sufficient allowance for intraspecific variation, none of the 
species are described, the only figures (of genitalia) are inadequate, and 
finally, she did not recognize five of the species now known from the 
United States; A. femoratus, A. tyleri, A. nasutus, A. peninsularis and A. 
leunculus. Despite these shortcomings, Sandhouse's revision has remained 
the best available reference on the genus for more than thirty years. More 
recent systematic treatments of Agapostemon are the contribution by 
Michener (1951) to the catalog of North American Hymenoptera; and the 
key, descriptions and figures for the species in the eastern United States 
(Mitchell, 1960). 

EXCLUDED SPECIES 

The following is a list of species which, although described in or sub- 
sequently transferred to Agapostemon, are no longer considered to belong 
to this genus. When possible the genus to which they belong has been 
indicated. Some of the species are synonyms but listing of synonymies must 
be postponed until Paragapostemon and Pseudagapostemon can be revised. 

In 1918(a) Cockerell reduced Pseudagapostemon to a subgenus of 
Agapostemon but this classification has never won acceptance. The only 
species mentioned by Cockerell [Agapostemon (Pseudagapostemon) 
xanthorhinus Cockerell, Halictus citricornis Vachal, Pseudagapostemon 
paulista Schrottky, and Pseudagapostemon nasua Schrottky] have been 
omitted from the following list. 

Augochloropsis 

Agapostemon caeruleus Ashmead 1890 was placed in Augochloropsis by 
Titus (1901). 



Revision of the Bee Genus Agapostemon 443 

Paragapostemon 

Notnia caelestina Westwood 1875, placed in Agapostemon by Cockerell 
(1910a), was placed in Paragapostemon by Moure (1964). 

Agapostemon bruneri Crawford 1901 was placed in Paragapostemon 
by Moure (1964). 

Halictus (Agapostemon) sicheli Vachal 1901 was placed in Paraga- 
postemon by Moure (1964). 

Nomia tacita Cameron 1902, placed in Agapostemon by Cockerell 
(1910a), was placed in Paragapostemon by Moure (1964). 

Nomia cillaba Cameron 1902, placed in Agapostemon by Cockerell 
(1910a), was placed in Paragapostemon by Moure (1964). 

Rttizantheda 

Halictus emarginatus Spinola 1851, placed in Agapostemon by Cockerell 
(1905), is a synonym of the type species (Halictus proximus Spinola 
1851) o£ Rttizantheda Moure 1964. 

Halictus mtttabilis Spinola 1851, placed in Agapostemon by Schrottky 
(1903), became the type species of Rttizantheda (Ruizanthedella) 
Moure 1964. 

Halictus placid us Smith 1879 was placed in Agapostemon by Cockerell 
(1905), but Moure (in litt.) places it in Rttizantheda. 
Pseudagapostemon 

Agapostemon arenaritts Schrottky 1902(b) became the type species of 
Pseudagapostemon Schrottky 1909(a). 

Agapostemon aeneus Schrottky 1902(a) was placed in Pseudagapostemon 
by Schrottky (1909b). 

Agapostemon arechavaletae Schrottky 1908 was placed in Pseudaga- 
postemon by Schrottky (1909b). 

Agapostemon bonaerensis Schrottky 1908 was placed in Pseudaga- 
postemon by Schrottky (1909b). 

Halictus (Agapostemon) pissisi Vachal 1903 was placed in Pseudaga- 
postemon by Vachal (1904). 

Halictus (Agapostemon) divaricattts Vachal 1903 was placed in Pseud- 
agapostemon by Vachal (1904). 

Agapostemon olivaceo-splendens Strand 1910 was placed in Pseudaga- 
postemon by Moure (1947). 

Agapostemon zosteronedys Moure 1940 belongs in Pseudagapostemon 
divaricattts (Vachal) as indicated by Moure (footnote in Michener 
and Lange, 1958). 
Species Incertae Sedis 

Halictus bruchianus Schrottky 1908 was placed in Agapostemon by 
Schrottky (1913). Moure (in litt.) has not seen the type but believes 
it should be placed in Corynttra or Rttizantheda (probably the latter). 



444 The University Science Bulletin 

GENERIC DIAGNOSIS 

No one character or set of characters was found to be both necessary 
and sufficient to distinguish all species of Agapostemon from all species in 
other halictine genera. However, the genus may be recognized by a syn- 
drome of characters. This set of characters is common to most species of 
Agapostemon but not all of its characters are present in all of the species. 

The character which is unique to Agapostemon and which best separates 
this genus from other halictine genera is the carina which in most species 
entirely surrounds the posterior surface of the propodeum of both sexes. 
Also unique among halictines is the striking contrast between the non- 
metallic coloration of the metasoma and the metallic coloration of the head 
and mesosoma of most female and nearly all male Agapostemon. 

Female Agapostemon may be distinguished from females of most other 
halictine genera by the three or four (sometimes as many as seven) large 
spatulate teeth on the posterior hind tibial spur and by the parallel con- 
tiguous carinae extending postero-dorsally from the antero-ventral margin 
of the gena. Male Agapostemon may be distinguished from males of the 
augochlorine genera and the genera in the Halictus-Lasioglossum group by 
the fusion of the first two tarsomeres of the hind tarsus. Unlike those of 
almost all other halictine genera, the hind femora and, to a lesser extent, 
the hind tibiae of many male Agapostemon are conspicuously swollen. The 
only other genera with similarly modified legs have conspicuously hairy 
eyes in contrast to the glabrous or nearly glabrous eyes of Agapostemon. 

INFRAGENERIC STRUCTURE 

While it seems premature to speculate at length on the phylogeny of 
the species of Agapostemon, certain discrete groups of species can be 
recognized on the basis of such features as genitalia, pronotum, legs, 
metasomal sterna and color pattern. These species groups might have been 
recognized as subgenera, but to be consistent I would have been forced to 
recognize an unacceptable number of small or monotypic subgenera. Sub- 
species are not recognized because there is little biological or utilitarian 
justification for such in this genus. Of course, intra-specific variation is 
described wherever encountered. 

In the following list the name of the first member (chosen arbitrarily) 
of each species group has not been indented and will serve to identify the 
group in subsequent discussion (e.g., the species in the splendens group 
are A. splendens, A. texanus and A. angelicas). 



NORTH 
AMERICAN 



SOUTH 
AMERICAN 

(except atrocaenileus) 



I{ohliellus 

centratus 

poeyi 
insularis 
jamaicensis 
vieqitesensis 
colit m bi 
ochromops 
sapphirinus 
cyan ens 
aenigma 

virididus 
obscuratus 
hispaniolicits 
c u ben sis 
alayoi 
swainsonae 



splendens 
texanus 
angelicus 

radiatus* 
cockerelli 
femoratus* 

virescens* 
tyleri 
color -ad in us* 

melliventris 
peninsularis 
mexicanus 



rhopiilocera 
erebus 



as cms 
leunculus 

n as ut us 

atrocaenileus 
semimelleus* 
chapadensis* 

intermedins 



heterurus 
inca 
mourei 
boliviensis 
lanosus 



WEST 
INDIAN 



MESO-AMERICAN 
(except those 
with asterisk) 



446 The University Science Bulletin 

ZOOGEOGRAPHY 

Although more than 50,000 specimens were examined in the course of 
this study, I was hampered by inadequate collections from the West Indies 
as well as Central and South America. While it seems unlikely that any new 
North American species will be discovered, it is likely that new species 
remain to be found in the West Indies, Central America and the Andean 
region of South America. 

It is difficult to comprehend what factors influence the distribution of 
species of Agapostemon. Only in the correlation between the distribution 
of A. splendens and that of moist sandy soils in the eastern United States 
and eastern Mexico is there evidence of a causal relation between the 
presence of an ecological parameter and the presence of the bees (Roberts, 
1969). 

Some species, such as A. coloradinus, are relatively restricted in range 
while others are widespread, A. texanus occurring from Canada to Costa 
Rica and Boston to San Francisco. In the United States alone, A. texanus 
occurs in 70 of 116 plant communities (Appendix A). A. angelicas occurs 
from below sea level in Death Valley, California to 12,000 ft. (3,658 m) on 
Mt. Evans in Colorado (timberline 11,700 ft.). In view of the apparent 
ecological plasticity among most species, it is not surprising that their dis- 
tributions broadly overlap. 

North American Species. There are four North American species 
groups. Each of these groups is composed of two very similar species and 
a third species which, although obviously closely related, stands somewhat 
apart phenetically. In each group but one, the two most similar species are 
found west of the 95th meridian and the third species is found north and 
east of the others. There is some sympatry within each species group, but 
the center of distribution is different for each species within a group (Fig. 
228). 

The simplest explanation for the occurrence of these similar geographic 
and cladistic patterns in each of the North American species groups is to 
assume: (1) North America was originally occupied by four species; (2) 
each species was bisected into southeastern and southwestern populations 
at the time of Pleistocene glaciation; and (3) the southwestern populations 
were more recently subdivided (possibly on the Pacific and Gulf coasts of 
Mexico) when forced to migrate farther south by still further cooling. 
Presumably this pattern of evolution is not evident among the Meso- 
American and West Indian species groups because they were too far south 
to have been significantly displaced by the climatic changes accompanying 
Pleistocene glaciation. 

West Indian Species. The West Indian species are interesting in that 
they are not found on the mainland (with the doubtful exception of A. 



Revision of the Bee Genus Agapostemon 447 

aenigma) and are not closely related to the mainland species. This is 
somewhat surprising in view of their relative proximity to the Florida and 
Yucatan peninsulas. Also surprising is their absence in the lesser Antilles; 
yet their occurrence throughout the Bahama Islands and Greater Antilles 
constitutes prima-facie evidence of high vagility. 

Although it is extremely improbable that the bees fly of their own 
accord between islands, it does seem likely that they are occasionally swept 
up by hurricanes and deposited on other islands many miles distant. Mem- 
bers of the poeyi and viridulus groups are widespread (cf. map, Fig. 1). 
As there is very little discernible correlation of geographic factors with the 
phenotype among members of each of these phenetically homogeneous 
species groups, I am inclined to believe that, relative to their rates of evolu- 
tion, their rates of dispersal have been rapid. 

In studying the poeyi group I vacillated between considering all of the 
island populations as conspecific and considering each population as a 
separate species. A. poeyi and A. viequesensis, occurring in Cuba and Puerto 
Rico respectively, have long been considered as distinct species which could 
easily be differentiated morphologically. However, as morphologically in- 
termediate forms exist on the Bahama Islands, it is tempting to classify the 
entire group as a single highly variable species. However, phenetically 
typical populations of A. poeyi and A. viequesensis are sympatric on New 
Providence Island, and there is no evidence of hybridization. The phenetic 
homogeneity within each of the phenetically different populations of this 
complex on other islands leads me to believe that gene flow between these 
allopatric populations is also inconsequential. Thus it seems that in the 
poeyi group the rate of speciation, or creation of clades, is high relative to 
the rate of evolution, or shift in gene frequency. Although not as large 
and widespread, the viridulus group similarly appears to be speciating rela- 
tively rapidly. My decision to regard the A. poeyi group as an Arten\reis 
rather than a Rassen/^reis is based on scanty evidence. However, the de- 
cision is a taxonomic necessity which in no way effects the biological at- 
tributes of the organisms in question. 

The species A. l{ohliellus and A. centratus do not seem closely related to 
any other species of Agapostemon. Rare species, they may be restricted to 
a particular ecological factor found only on the island of Hispaniola. Such 
a severe ecological limitation could explain their absence on the other islands. 

Meso-American Species. The distributions and relationships of the 
species occurring in this region are difficult to interpret. Some of the species 
such as A. mexicanus or A. tyleri clearly belong to North American groups 
although A. mexicanus is found in northwestern Mexico and A. tyleri 
ranges southward on the central plateau to the vicinity of Mexico City. 

A. rhopalocera is a rare species known only from males, not obviously 



448 The University Science Bulletin 

related to other species and restricted to the vicinity of Mount Orizaba in 
Vera Cruz. A. erebns is known from a single female collected in El Salva- 
dor, but is obviously closely related to A. leunculus which ranges from the 
southern tip of Texas to eastern Ecuador, and also to a third species, A. 
ascitis, known from two females collected in Colombia. 

Males of A. nasutus are unlike those of any other species inasmuch as 
their genitalia and clypeal region are both highly modified. They also differ 
from males of other species in less striking features such as the lateral tufts 
of pubescence on the 6th metasomal sternum. Despite these obvious peculi- 
arities of males, females of A. nasutus are somewhat similar to those of the 
erebus group (differing principally in the shape of the pronotum). The 
commonest species throughout most of its range, A. nasutus is found from 
the southern tip of Texas to northern South America, as far east as Trinidad, 
and west of the Andes as far south as lea, Peru. 

A. atrocaeruleus has been found only in Costa Rica but is closely related 
to a pair of species found primarily south of the Amazon Basin but reaching 
as far north as Ecuador. A. intermedins is found from Costa Rica to Tingo 
Maria, Peru and obviously links the atrocaeruleus group with the Andean 
heterurus group. 

In summary, most of the Meso-American species of Agapostemon belong 
in North or South American species groups. Only the very widespread and 
abundant A. nasutus and the very restricted and rare A. rhopalocera have no 
apparent affinities with either North or South American species, or each 
other. 

South American Species. Some of the species occurring in South Amer- 
ica such as A. nasutus, A. intermedins and A. ascitis are restricted to the 
northwestern margin of the continent. Apparently these species, which are 
recently evolved or recently arrived via Central America (e.g., A. nasutus), 
have found further access to the continent blocked by the Andes to the east, 
the Atacama Desert to the south, and the forests of the Orinoco Basin to the 
north. 

Agapostemon chapadensis and A. semtmelleus are the only species widely 
distributed in South America. Both are found in the campos, or subtropical 
grasslands, of southern Brasil, Paraguay and northern Argentina. In addition 
isolated populations of A. semimelleus have been collected from the head- 
waters of the Marafion, Huallaga and Cauca rivers. Presumably, these species 
were able to ascend the valleys at some time in the past when the South 
American grasslands were more extensive than today. Surprisingly, A. 
semimelleus has been collected at the head of the Cauca Valley in Colombia. 
I cannot explain this occurrence as the Cauca River drains into the Caribbean 
Sea. Specimens of A. semimelleus have been collected from localities as near 
as the headwaters of the Putumayo River, but the Cauca and Putumayo 



Revision of the Bee Genus Agapostemon 449 

rivers are separated by at least 200 kilometers of mountains (Cordillera Cen- 
tral) rising as high as 5755 m. The closest relative of A. semimelleus and A. 
chapadensis is A. atrocaerideas which has been collected only on the central 
plateau of Costa Rica, nearly 1000 km from the headwaters of the Cauca 
River. 

The most aberrant species group in Agapostemon includes A. heterurus, 
A. inca, A. boliviensis, A. lanosus and A. mourei. These species are very 
closely related and are restricted to the eastern margins of the central Andes. 
As many of the valleys in this region are both ecologically isolated and poorly 
collected, it seems reasonable to predict that more species in this group will 
be discovered. 

A. intermedins is morphologically intermediate between the A. heterurus 
group and the A. atrocaeruleus group. A. intermedins ranges from central 
Costa Rica, where it is sympatric with A. atrocaeruleus, to northwestern 
South America. It has also been collected from Tingo Maria at the head of 
the Huallaga Valley where it is sympatric with A. lanosus. Possibly A. 
intermedins is part of an Arten\reis running from Costa Rica down to the 
grasslands of Brasil (the A. atrocaeruleus group) and down to the eastern 
margins of the Andes (the A. heterurus group). 

METHODS 

Terminology. The terminology is essentially that utilized by Michener 
(1944, 1965). However, the following terms have been proposed since no 
appropriate ones existed for these features. 

The propodeal shield is the flat or slightly concave area on the posterior 
vertical surface of the propodeum usually delimited by a conspicuous propo- 
deal carina (Fig. 25). Male genitalia often bear an apical stylus, medial plate 
and basal stylus on the mesal surface of the gonostylus (Fig. 180). Although 
often reduced and covered by long bristles, a ventral lobe (Fig. 223) is always 
present on the ventral surface of the gonocoxite. 

The specific importance of the maculations of male legs has necessitated 
a more precise and morphologically consistent system for designating various 
portions of the legs than that used by Michener and others. The legs of 
bees, like those of most insects, may move anteriorly and posteriorly, and may 
be flexed ventrally and extended laterally. The middle legs, when extended, 
are nearly perpendicular to the longitudinal axis of the body (presumably 
the "primitive" condition in the Arthropoda). The fore legs and hind legs, 
however, are directed respectively forward and rearward to a greater or 
lesser degree. Owing to the rearward orientation of the hind legs of bees 
(especially in pinned specimens and in live bees in flight), that which is 
designated by previous authors as the "inner" surface of the hind legs is 
not homologous with what they term the "inner" surface of the fore legs. 



450 The University Science Bulletin 

Therefore, I propose a terminology consistent with the presumed serial 
homology of the parts of the three pairs of legs rather than with the positions 
in which they are borne. Although this system may at first seem strange 
and cumbersome, it is economical in that one can refer, for example, to 
serially homologous maculations as being on the posterior surfaces of all 
three pairs of legs. With the system used by other authors one could not 
refer to such maculations as being on the "inner" surfaces of all three 
pairs of legs. The disadvantage of the proposed system is that the "dorsal" 
quadrant of the tarsus may be visible only from below (due to the flexed 
position of the leg), and the "anterior" surface of the hind leg only from 
the side. 

In the following descriptions, the legs are considered as extended at right 
angles to a sagittal plane through the body of the bee. The hypothetical 
leg segment is regarded as cylindrical, its cross-section divided by perpen- 
dicular diagonals into dorsal, ventral, anterior and posterior quadrants. 

Description. To conserve space, the reader is often referred to the de- 
scription of another species or sometimes the opposite sex. All interspecific 
comparisons are between members of the same sex unless otherwise specified. 

Because sexual dimorphism is so pronounced in Agapostemon, characters 
used to differentiate males (or females) of two closely related species are not 
always present on members of the opposite sex. In some instances members 
of only one sex can be differentiated morphologically. In other cases the 
morphological differences between species are so subtle as to render identi- 
fication difficult regardless of sex. For these reasons distributions have been 
used (especially in the keys) wherever morphological differentiation was 
difficult or impossible. Of course, no species is recognized solely on the basis 
of its distribution. The reader must be cautious in relying on distributional 
differentia because the ranges of species are subject to change without notice. 

If two species are very similar then only one is described and the other 
is described as identical but for the differentiating characters. Relative size 
is presented as proportions and absolute size may be derived from scale 
drawings. For each character, the drawings are to the same scale to 
facilitate size comparisons between species. 

Synonymies have been kept as brief as possible and no references to 
synonyms have been made unless they involve nomenclatural changes of 
specific epithets. An effort was made to locate and examine the primary 
types of all nominal species and subspecies and, whenever possible, their 
location has been noted in the synonymy. In quotations of label data on 
the types of new species, a single slash indicates the break between lines 
on a label and a double slash the break between labels on a pin. 



Revision of the Bee Genus Agapostemon 451 

KEY TO SPECIES OF THE UNITED STATES 
AND CANADA 

I. Female; 10 flagellomeres; scopa on hind leg 2 

Male; 11 flagellomeres; scopa absent 14 

2.( 1 ) Metasomal terga bright metallic green to blue, concolorous with 

head and mesosoma 3 

Metasomal terga black to pale amber, not concolorous with 
metallic head and mesosoma 7 

3.( 2 ) Wings almost hyaline, only slightly darkened on distal margins; 

mandible yellow basally 4 

Wings transparent brown, distal margins conspicuously dark- 
ened; mandible usually amber basally (rarely yellow) splendens 

4.( 3 ) Mesoscutum with numerous fine punctures interspersed with 
fewer distinctly larger and deeper punctures; punctuation often 

so fine as to leave mesoscutum shiny (Figs. 23, 24) 

texanus and angelicus* 

Mesoscutum coarsely punctate or rugose, lacking punctures of 

two distinct sizes and never shiny 5 

5.( 4 ) Mesoscutum coarsely rugose at least on inner borders of 

parapsidal lines, if not rugose throughout jemoratus 

Mesoscutum between parapsidal lines coarsely punctate, not 
rugose 6 

6.( 5 ) Common in eastern half of U.S., becoming rare in the Great 

Plains (Fig. 18) radiatus 

Common on Mexican Plateau and in arid southwestern U.S., rare 

along eastern border of Rocky Mountains (Fig. 2) cockfrelli 

7.( 2 ) Posterior lobe of pronotum with pale yellow or creamy spot at 
apex (Fig. 134); clypeus with yellow transverse band (Fig. 42) 

nastttus 

Posterior lobe of pronotum metallic at apex, never creamy or 
yellow; clypeus with or without yellow transverse band 8 

8.( 7 ) Clypeus with transverse apical or subapical yellow band; 

metasoma black or pale amber 9 

Clypeus metallic with apical region dark brown to black; 
metasoma always black 10 

9.( 8 ) Pronotum with single conspicuous sharp carina extending 
postero-ventrally from lateral angle (Fig. 135); metasomal terga 
black with narrow basal bands of white tomentum; scape dark 

brown to black, lacking yellow leunculus 

Pronotum without conspicuous sharp carina extending postero- 
ventrally from lateral angle (may have several small carinulae); 
metasomal terga pale amber to black, with white tomentum on 

basal halves; scape sometimes marked with yellow 12 

10.(8) Genal ridges coarse (2-3 per 0.25 mm) (Fig. 28); mandible 

usually yellow basally virescens 

Genal ridges fine (5-6 per 0.25 mm); mandible amber to brown- 
black basally 1 1 

* The females of A. texanus and A. angelicus cannot be reliably separated on the basis of 
their morphology; however, females occurring well outside the range of A. angelicus males are 
probably A. texanus (see map, Fig. 20). 



452 The University Science Bulletin 

11.(10) Wings transparent brown; white tomentum lacking on anterior 
part of first metasomal tergum; protuberance above clypeus 

without large, central, shiny area coloradinus 

Wings hyaline, not brown; white tomentum on anterior part of 

first metasomal tergum 13 

12.( 9 ) Dorsal area of propodeum moderately to coarsely rugose; scape 
often marked with yellow; metasomal terga amber to black or 

black with amber anteriorly on tergum 1 melliventris 

Dorsal area of propodeum finely rugose to finely rugulose; scape 
never with yellow markings; metasomal terga always entirely 
black peninsularis 

13.(11) Supraclypeal protuberance with smooth shiny central area with 
scattered punctures; found only in Arizona, New Mexico and 

on the Mexican Plateau (Fig. 27) tyleri 

Supraclypeal protuberance with weakly and transversely rugose 
central area; occurs around Gulf of California mexicanus** 

14. ( 1 ) Clypeus conspicuously concave ventrally, more than three times 

as broad as long (Figs. 86, 87) nasutus 

Clypeus normal, little broader than long 15 

15.(14) First metasomal tergum yellow to very pale amber basally, not 
brown or black, and hind basitarsus always slender and simple 
(Fig. 155); hind leg yellow with small brown or black area 

apically on femur and basally on tibia melliventris 

First metasomal tergum brown to black basally (if pale amber, 
then hind basitarsus swollen and with basal ridge and apical 
groove); hind leg usually with dark brown to black stripe on tibia 16 

16.(15) Hind femur without tooth; slender 25 

Hind femur with subapical tooth postero-ventrally; often con- 
spicuously inflated 17 

17.(16) Wings transparent brown, conspicuously darkened at distal 
margins; hind basitarsus with crest of basal ridge grooved 

(Fig. 166) splendens 

Wings hyaline or nearly so, not conspicuously darkened at distal 
margins; hind basitarsus with basal ridge (if present) not grooved 18 

18.(17) Metasoma with last two visible sterna (5-6) dark brown to 
brown-black, without yellow maculations; sterna 2-4 with yellow 

(if present) restricted to basal margins 19 

Metasoma with extensive yellow maculations on last two visible 
sterna (5-6) as well as on sterna 2-4 21 

19.(18) Hind femur lacking brown to black stripe on posterior surface 

( Fig. 158) virescens 

Hind femur with large brown to black stripe covering most or 

all of posterior surface 20 

20.(19) Brown to black streak less than % length of posterior surface of 

hind tibia (Fig. 160) coloradinus 

Brown to black streak extending entire length of posterior surface 

of hind tibia (Fig. 159) tyleri 

** Although not currently known to occur in the United States, this species may yet be 
found in the vicinity of San Diego or Yuma, thus it is included in this key. 



Revision of the Bee Genus Agapostemon 453 

21.(18) Metasomal sternum 4 (antepenultimate visible) with low trans- 
verse ridge not quite reaching posterior margin laterally; 

metasomal tergum 4 usually without conspicuous metallic tints 22 

Metasomal sternum 4 (antepenultimate visible) with low trans- 
verse ridge reaching posterior margin laterally; metasomal 
tergum 4 nearly always with conspicuous metallic tints 24 

22.(21) Hind basitarsus lacking apical groove and basal ridge (Fig. 

161) radiatus 

Hind basitarsus with apical groove and basal ridge 23 

23.(22) Hind basitarsus with inconspicuous basal ridge and slender apical 

groove (Fig. 162) coc\erelh 

Hind basitarsus with very large, conspicuous, slightly sinuate 

basal ridge and broad, conspicuous apical groove (Fig. 163) .... jemoratus 

24.(21) Base of apical stylus of gonostylus slightly inflated (Fig. 181); 
brown to black stripe on posterior surface of hind tibia but 

never on anterior surface (Fig. 164) angelicus 

Base of apical stylus of gonostylus not inflated (Fig. 180); brown 
to black stripe anteriorly on hind tibia (Fig. 165), or, if lacking, 
then without black stripe on posterior surface texanus 

25.(16) Posterior surface of hind tibia largely yellow (Fig. 143) leunculus 

Posterior surface of hind tibia largely brown or black 26 

26.(25) 2-4 large submarginal bristles on each side of metasomal 

sternum 4 peninsularis 

14-20 large, evenly spaced submarginal bristles on metasomal 
sternum 4 mexicanus* 

KEY TO MESO-AMERICAN** SPECIES 

1. Males; 11 flagellomeres; scopa absent 2 

Females; 10 flagellomeres; scopa on hind legs 15 

2.( 1 ) Antenna normal 3 

Antenna long, filamentous, and with apical flagellomere flattened 

(Fig. 131 ) rhopalocera 

3.( 2 ) Clypeus normal, little longer than broad 4 

Clypeus conspicuously concave ventrally, more than three times 

as broad as long (Figs. 86, 87) nasutus 

4.( 3 ) Metasoma banded with yellow and dark brown or black, lacking 

bands of white tomentum 5 

Metasoma pale amber, with bands of white tomentum centrums 

5.( 4 ) Sixth metasomal sternum flat or with inconspicuous medial ridge 6 

Sixth metasomal sternum with medial "button" (flattened process 
resembling human tongue and attached to sternum in much the 
same way as human tongue is attached to floor of mouth 
(Fig. 136) intermedins 

6.( 5 ) Hind femur with conspicuous subapical tooth postero-ventrally 7 

Hind femur lacking tooth 1 1 



* Although not currently known to occur in the United States, this species may yet be 
found in the vicinity of San Diego or Yuma, thus it is included in this key. 
** Mexico, Panama, and intervening countries. 



454 The University Science Bulletin 

7.( 6 ) Wings hyaline or nearly so, not conspicuously darkened at distal 
margins; hind basitarsus with basal ridge (if present) not 

grooved 8 

Wings transparent brown, conspicuously darkened at distal 
margins; hind basitarsus with crest of basal ridge grooved 

(Fig. 166) splendens 

8.( 7 ) Metasoma with extensive yellow maculations on sterna; hind 
tibia with brown or black posterior stripe (when present) not 

extending from base to apex 9 

Metasoma with sterna brown or black, lacking yellow macula- 
tions; hind tibia with broad brown or black posterior stripe 

extending from base to apex tyleri 

9.( 8 ) Metasomal sternum 4 (antepenultimate visible) with low trans- 
verse ridge reaching posterior margin laterally; metasomal 

tergum 4 nearly always with conspicuous metallic tints 10 

Metasomal sternum 4 (antepenultimate visible) with low trans- 
verse ridge not quite reaching posterior margin laterally; meta- 
somal tergum 4 without conspicuous metallic tints coc\erelh 

10. ( 9) Base of apical stylus of gonostylus slightly inflated (Fig. 181); 
brown to black stripe on posterior surface of hind tibia but 

never on anterior surface (Fig. 164) angelicas 

Base of apical stylus of gonostylus not inflated (Fig. 180); brown 
to black stripe anteriorly on hind tibia (Fig. 165) or, if lacking, 
then without black stripe on posterior surface texanus 

ll.( 6 ) Base of metasomal tergum 1 dark brown or black 12 

Base of metasomal tergum 1 yellow or pale amber melliventris 

12.(11) Hind leg with broad brown or black stripe covering most of 
posterior surface of tibia and extending from base to apex; 

similar stripe on femur sometimes broken centrally 13 

Hind leg largely yellow on posterior surface of tibia and femur 14 

13.(12) Metasomal sternum 4 with 2-4 large submarginal bristles on 

each side peninsularis 

Metasomal sternum 4 with 14-16 large, evenly spaced sub- 
marginal bristles mexicanus 

14.(12) Pronotum with very acute lateral angle; mesoscutum with con- 
spicuous flange laterally on anterior margin atrocaeruleus 

Pronotum with lateral angle rounded; mesoscutum lacking flange 

on anterior margin leunculus 

15.( 1 ) Metasomal terga bright metallic green to blue-green 16 

Metasomal terga pale amber to black, not metallic 19 

16.(15) Mandibles amber with metallic green spot basally; dorsal area 
of pronotum with very large, widely separated carinae extend- 
ing from anterior margin to propodeal carina aenigma* 

Mandibles amber or yellow, lacking metallic spot basally; pro- 
notum rugose dorsally 17 

17.(16) Wings almost hyaline, only slightly darkened on distal margins; 

mandibles yellow basally 18 

Wings transparent brown, distal margins conspicuously dark- 
ened; mandibles usually amber (sometimes yellow) basally .... splendens 

* Although labeled "Costa Rica" the two specimens of A. aenigma are closely related to 
West Indian species and may be mislabeled. 



Revision of the Bee Genus Agapostemon 455 

18.(17) Mesoscutum coarsely punctate to finely rugose, lacking punctures 

of two distinct sizes and never shiny coc\erelli 

Mesoscutum with numerous fine punctures interspersed with 
fewer distinctly larger and deeper punctures, punctation often 

so fine as to leave mesoscutum shiny (Figs. 23, 24) 

texanus and angelicas** 

19.(15) Lateral angle of pronotum acutely pointed; mesoscutum with 

conspicuous flange laterally on anterior margin 20 

Lateral angle rounded, not acutely pointed; mesoscutum lacking 
conspicuous flange on anterior margin 21 

20.(19) Head and mesosoma black (usually with inconspicuous dark 
blue tints and with yellow band on clypeus); metasoma black 

with bands of white tomentum atrocaeruleus 

Head and mesosoma metallic green or coppery; metasoma pale 
amber to black, with yellow integumental bands replacing usual 
bands of white tomentum intermedins 

21.(19) Posterior lobe of pronotum metallic at apex, never creamy or 

yellow; clypeus with or without yellow transverse band 22 

Posterior lobe of pronotum with creamy or yellow spot at apex; 
clypeus with yellow transverse band nasntus 

22.(21) Clypeus with transverse, subapical. yellow band; metasoma 

amber to black 23 

Clypeus black at apical margin, lacking yellow band; metasoma 
always black 25 

23.(22) Pronotum without single conspicuous sharp carina extending 
postero-ventrally from lateral angle ( may have several small 
carinulae); metasomal terga pale amber to black, with white 
tomentum on basal halves; scape dark brown to black, sometimes 

marked with yellow 24 

Pronotum with single conspicuous sharp carina extending 
postero-ventrally from lateral angle; metasomal terga with nar- 
row basal bands of white tomentum; scape dark brown to black, 
lacking yellow leunculus 

24.(23) Dorsal area of propodeum moderately to coarsely rugose; scape 
often marked with yellow; metasomal terga amber to black, or 

black with amber anteriorly on tergum 1 mellwentris 

Dorsal area of propodeum finely rugose to finely rugulose; scape 
never with yellow markings; metasomal terga always entirely 
black peninsularis 

25.(22) Head and mesosoma bright metallic green 26 

Head and mesosoma black with metallic purple tints erebus 

26.(25) Supraclypeal protuberance with smooth shiny central area and 
with scattered punctures; occurs in New Mexico, Arizona, and 

on the Mexican Plateau (Fig. 27) tyleri 

Supraclypeal protuberance with weakly and transversely rugose 
central area; occurs around the Gulf of California mexicanus 



s 



** The females of A. texanus and A. angelicits cannot be separated reliably on the basis 
of their morphology; however, females occurring well outside the range of A. angelicus males 
are probably A. texanus (see map, Fig. 20). 



456 The University Science Bulletin 

KEY TO WEST INDIAN SPECIES 

1. Males; 11 flagellomeres; scopa absent 2 

Females; 10 flagellomeres; scopa on hind leg 16 

2.( 1 ) Hind femur without tooth 3 

Hind femur with tooth on postero-ventral margin 4 

3.( 2 ) Lower portion of clypeus abruptly flattened, glabrous and im- 
punctate (Figs. 109-110); metasomal terga yellow with black 

bands; from Hispaniola, Cuba and Jamaica h^ohliellus 

Lower portion of clypeus normal, not abruptly flattened, 
glabrous (Figs. 121-122) or impunctate; metasomal terga honey- 
colored; from Haiti centratus 

4.( 2 ) Metasomal terga black or pale amber, never with metallic tints; 

clypeal region elongate 5 

Metasomal terga with yellow and brown to black bands, often 

with metallic tints; clypeal region normal 9 

5.( 4 ) Head and mesosoma bright metallic green to blue 6 

Head and mesosoma shiny brown-black, not metallic; from 

Cuba obscuratus 

6.( 5 ) Metasomal terga dark brown to black, pale bands of tomentum 

conspicuous 7 

Metasomal terga pale amber with brown bands, pale bands of 
tomentum inconspicuous or absent 8 

7.( 6 ) Sculpturing of mesosoma slightly finer than that of vindulus, 
contiguous punctures giving the mesoscutum a dark blue ap- 
pearance when viewed from above; from Hispaniola hispaniolicus 

Sculpturing of mesosoma slightly coarser than that of hispanioli- 
cus, subcontiguous punctures giving the mesoscutum slightly 
shiny green to blue reflections; from Cuba vindulus 

8.( 6 ) Basal ridge of hind basitarsus very broadly and deeply grooved 
(Fig. 147); mesoscutellum extremely shiny, only weakly sculp- 
tured; from Jamaica swainsonae 

Basal ridge of basitarsus very narrow and without broad, deep 
groove (Fig. 149); mesoscutellum weakly shiny with contiguous 
punctures; from Cuba cubensis 

9.( 4 ) Eye normal brown; pterostigma translucent amber to dark brown 10 

Eye and (or) underside of pterostigma (except for dark mar- 
gins) pale cream colored to opaque yellow 14 

10.( 9 ) Metasomal tergum 3 weakly to strongly metallic green to blue 

medially 11 

Metasomal tergum 3 without metallic green to blue medially 13 

11.(10) Metasomal tergum 1 with metallic green to blue tints postero- 

laterally 12 

Metasomal tergum 1 without metallic green to blue tints postero- 
laterally; from Hispaniola insidaris 

12.(11) Hind femur 60% as wide as long (Fig. 146); from Jamaica .. jamaicensis 
Hind femur less than 55%, as wide as long (Fig. 144); from 
Cuba, Hispaniola, New Providence I. and probably from Andros 
I. and Cat I poeyi 



Revision of the Bee Genus Agapostemon 



457 



13.(10) Metasomal terga 4 and 5 with conspicuous metallic green tints 

medially; from San Salvador I. (= Wading I.) columbi 

Metasomal terga 4 and 5 without conspicuous metallic green 
tints medially (may have faint metallic tints laterally); from 
Puerto Rico (including Mona I. and Vieques I.) and New 

Providence I viequesensis 

Pedicel, unlike scape or flagellum, pale amber to yellow below 15 

Pedicel, unlike scape or flagellum, dark brown below; from 

Crooked I cyaneus 

From Long I sapphirinus 

From Cat I., Rum Cay, Conception I., New Providence I. and 

Mariguana (=Mayaguana?) I ochromops 

Metasomal terga black to pale amber, not metallic 17 

Metasomal terga largely metallic green to blue to purple 22 

Lower portion of clypeus brown or black, metasomal terga pale 

amber to black, lacking yellow bands 18 

Lower portion of clypeus with broad yellow band; metasomal 

terga with yellow bands; from Hispaniola, Cuba and Jamaica .. \ohliellus 

Metasomal terga uniformly brown-black with narrow, white 

bands of tomentum 19 

Metasomal terga 1 and 2 honey-colored becoming brown-black 

on posterior terga; from Hispaniola alayoi 

Head and mesosoma metallic green to blue 20 

Head and mesosoma shiny brown-black, not metallic; from 

Cuba obscuratus 

More than 50% of interocular area above antennal sockets and 

below median ocellus rugose to rugulose 21 

More than 50% of interocular area above antennal sockets and 
below median ocellus with fine, deep, contiguous punctures; 

from Hispaniola hispamolicus 

Mandible with metallic green tints basally, about as dark 

ferruginous-brown basally as apically; from Cuba viridulus 

Mandible without metallic green tints basally, much darker 

apically than basally; from Jamaica swainsonae 

Metallic coloration dark blue or dark purple 23 

Metallic coloration green 24 

Gena with fewer than 10 extremely coarse carinae; tegula and 

legs with some pale amber areas; from Long I sapphirinus 

Gena with many more than 10 fine carinae; tegula and legs 

dark brown; from Crooked I cyaneus 

Eye normal brown; pterostigma translucent amber to dark brown 25 

Eye or underside of pterostigma (usually both) pale cream- 
colored to opaque yellow; from Cat I., Rum Cay, Conception I., 

New Providence I. and Mariguana (=rMayaguana?) I ochromops 

Mandible with metallic green spot basally 26 

Mandible lacking metallic green spot basally; from Puerto Rico 
(including Mona I. and Vieques I.) and New Providence I. .. viequesensis 

Metasomal sterna 3 and 4 with conspicuous medial metallic 

green tints 27 

Metasomal sterna 3 and 4 brown, without medial metallic green 

tints 28 



458 The University Science Bulletin 

27.(26) Mesoscutum rugose laterally along parapsidal line, becoming 
coarsely rugose anteriorly, punctate centrally and posteriorly; 

from Cuba, Hispaniola, New Providence I., Cat I. and Andros I poeyi 

Mesoscutum finely punctate, with extreme antero-lateral margin 
rugulose; from Hispaniola insularis 

28.(26) Scopal hairs on femur and hairs on postero-ventral surface of 

hind tibia golden; from San Salvador I. ( = Watling I.) columbi 

Scopal hairs on femur and hairs on postero-ventral surface of 

tibia white; from Costa Rica aenigma* 

KEY TO SOUTH AMERICAN SPECIES 

1. Female; 10 flagellomeres; scopa on hind leg 2 

Male; 11 flagellomeres; scopa absent 10 

2.( 1 ) Mesoscutum rounded on anterior margin; lateral angle of 

pronotum inconspicuous and not projecting antero-laterally 3 

Mesoscutum with acute upturned flange on anterior margin; 
lateral angle of pronotum conspicuous and projecting antero- 
laterally 5 

3.( 2 ) Pronotum with creamy or yellow spot at apex of posterior lobe 
and without carina extending postero-ventrally from lateral angle 

(Fig. 134) nasutus 

Pronotum without creamy or yellow spot at apex of posterior lobe 
and with conspicuous sharp carina extending postero-ventrally 
from lateral angle (Fig. 135) 4 

4.( 3 ) Metasoma black; lateral angle of pronotum rounded; 3 spatulate 

teeth on posterior hind tibial spur leunculus 

Metasoma amber; lateral angle of pronotum angular; 5 (rarely 

4?) spatulate teeth on posterior hind tibial spur ascius 

5.( 2 ) Metasomal terga 2 and 3 amber or black with yellow basally 

(best seen in posterior view) 6 

Metasomal terga 2 and 3 amber or black but lacking yellow 
semimelleus or chapadensis* 

6.( 5 ) Head and mesosoma dull coppery, metallic greenish-black or 
bluish-black; mesoscutal pubescence dense and woolly; restricted 

to western South America 7 

Head and mesosoma usually bright metallic green (may be 
coppery in Central America); mesoscutal pubescence not espe- 
cially dense and finely branched; Central and South America 
intermedins 

7.( 6 ) Propodeum metallic blue-black or green-black dorsally; clypeus 

without yellow maculations heternms 

Propodeum metallic coppery dorsally; clypeus usually with 
yellow maculations 8 



* Agapostcmon aenigma is included in this key because its similarity to West Indian species 
casts doubt on the correctness of its "Costa Rica" label. A specimen with identical label data 
belongs to the Haitian species A. centratus. 

* The females of these species cannot be distinguished reliably (cf. A. semimelleus, 
Variation). 



Revision of the Bee Genus Agapostemon 459 

8.( 7 ) Propodeum with coarse parallel striae dorso-laterally; clypeus 

without yellow maculations lanosus 

Propodeum coarsely rugose dorso-laterally; clypeus with yellow 
maculations 9 

9.( 8 ) Posterior half of each metasomal tergum pale amber, prepygidial 

fimbria brown to black mourei 

Posterior half of metasomal tergum 1 and 2 dark amber, tergum 
3 brown, terga 4 and 5 and prepygidial fimbria brown-black .. bolwiensis 
10. ( 1 ) Last visible (6th) metasomal sternum with medial, button-like 

protrusion basally (Figs. 136-140) 14 

Last visible (6th) metasomal sternum without button-like 
protrusion 11 

11.(10) Clypeus normal, little longer than broad 12 

Clypeus conspicuously concave ventrally, more than three times 

as broad as long (Figs. 85-87) nasutus 

12.(11) Lateral angle of pronotum prominent, acute and projecting 
antero-laterally; first metasomal tergum dull, with large, deep, 

subcontiguous punctures 13 

Lateral angle of pronotum rounded; first metasomal tergum 

shiny, with small, shallow, scattered punctures leunculus 

13.(12) Hind femur with brown on posterior extending from base to 
apex (Fig. 172); gonostylus with distal stylus about twice as 

long as width of medial plate (Fig. 173) chapadensis 

Hind femur with brown on posterior limited to base and apex 
(Fig. 171); gonostylus with distal stylus little longer than width 
of medial plate (Fig. 174) semimelleus 

14.(10) Hind femur less than twice as long as broad; hind basitarsus ex- 
tensively marked with brown or black; basal ridge and apical 
groove on hind basitarsus large and conspicuous (Figs. 167- 

170) 15 

Hind femur more than twice as long as broad; hind basitarsus 
yellow, only rarely marked with brown; basal ridge and apical 
groove on hind basitarsus reduced and inconspicuous (Fig. 
151) intermedins 

15.(14) Antepenultimate visible sternite (4th) without conspicuously 

long, contiguous, hooked setae laterally on posterior margin 16 

Antepenultimate visible sternite (4th) with two or three con- 
tiguous, long (about twice length of adjacent setae), thickened, 
hooked setae laterally on posterior margin lanosus 

16.(15) Last visible sternite (6th) with "button" about l A as wide as 
sternite (Figs. 137, 139); basitarsus relatively broad (Figs. 

167, 170) 17 

Last visible sternite (6th) with "button" about l / 2 as wide as 
sternite (Fig. 140); basitarsus relatively slender (Fig. 168) inca 

17.(16) Tegula pale amber with yellow lunule anteriorly; medial plate 

of gonostylus large (Fig. 176) mourei 

Tegula dark ferruginous, without yellow lunule anteriorly; 
medial plate of gonostylus small and rounded (Fig. 177A) bolwiensis 



460 The University Science Bulletin 

DESCRIPTIONS, SYNONYMIES AND DISTRIBUTIONS 
Agapostemon aenigma n. sp. 

The name of this species refers to its mysterious affinity with certain 
West Indian species. 

Agapostemon aenigma is known from two female specimens labeled 
"Costa Rica" in the same handwriting and type of label as two males of 
the Haitian A. centratus. This species appears closely related to some West 
Indian species and may be mis-labeled as are the similarly labeled specimens 
of A. centratus. The holotype and paratype are in the Academy of Natural 
Sciences, Philadelphia. 

Diagnosis. The female may be distinguished from other green Central 
American Agapostemon by the coarse parallel carinae on the dorsal surface 
of its propodeum. Among West Indian females with bright metallic green 
to blue metasomas, A. aenigma may be distinguished from females of 
A. ochromops by its lack of milky to yellowish eyes and pterostigma; from 
females of A. viequesensis by the basal metallic spot on its mandible; from 
females of A. cohtmbi by the white (not golden) pubescence on the under- 
side of its hind femur and tibia; and from A. poeyi and A. insularis by its 
lack of strong metallic tints medially on sterna 3 and 4. This species is very 
close to A. insularis and the two may eventually prove to be conspecific. 

Description 

female 

General coloration of head, mesosoma and metasoma bright metallic 
greenish blue. Head: pubescence as in A. viequesensis. (1) La brum as 
in A. hispaniolicus. (2-4) Clypeus, interocular area and vertex as in A. 
viequesensis. (5) Gena much more coarsely carinate than in A. viequesensis 
but not quite as coarse as in A. poeyi. (6) Malar area absent. (7-8) 
Mandible and antenna as in A. poeyi. Mesosoma: pubescence as in 
A. viequesensis but slightly more fuliginous on mesonotum and metanotum. 
(9) Pronotum as in A. viequesensis but with lateral angle and posterior 
lobe more angular, and with fewer and coarser (not as coarse as in A. poeyi) 
horizontal carinae postero-laterally. (10-12) Mesoscutum, mesoscutellum 
and metanotum as in A. viequesensis. (13-14) Mesepisternum and metepi- 
sternum as in A. viequesensis but with sculpturing very slightly coarser. 
(15) Propodeum as in A. poeyi but with finer sculpturing (not quite as 
fine as in A. viequesensis). (16-17) Wing and tegula as in A. viequesensis 
but with tegula darker with metallic tints more extensive. (18-20) Fore, 
middle and hind legs dark brown, coxae with strong metallic tints. 
Pubescence dark amber to brown, scopal hairs white. Metasoma: as 
in A. viequesensis but with faint metallic tints laterally on sterna 3 and 
4 of paratype (cf. A. insularis 5 ) . 



Revision of the Bee Genus Agapostemon 461 

Agapostemon alayoi n. sp. 

This species is named after Dr. Pastor Alayo D. of La Habana, Cuba, 
who has contributed significantly to our knowledge of Antillean bees. 

I have seen a single specimen labeled "Ace. 1007-25/Port au Prince,/ 
Haiti., 1925//G. N. Wolcott/Coll." The type is in the United States Na- 
tional Museum, Washington, D.C. 

Diagnosis. The female may be distinguished from females of A. 
Itphliellus by the lack of yellow on its clypeus. No other West Indian 
species has females with metasomal terga 1-2 pale amber. This species is 
phenetically closest to A. viridiilns and A. hispaniolicus. It differs from 
both in the coloration of its metasoma, in having a more densely punctate 
mesoscutellum and in being larger than females of A. hispaniolicus. 

Description 

female 

General coloration of head and mesosoma metallic greenish blue, 
metasoma honey-colored anteriorly, brown-black posteriorly. Head: as 
in A. vmduliis but metallic blue, with very dark fuliginous hair on 
interocular area and vertex. Mesosoma: as in A. viridiilns but metallic 
blue, with slightly finer, deeper and more numerous punctures on meso- 
scutum and mesoscutellum. Mesoscutellum not shiny as in A. viridiilns. 
Sculpturing of mesepisternum, metepisternum and propodeum slightly 
deeper than in A. viridiilns. Pubescence of mesoscutum, mesoscutellum and 
metanotum much darker fuliginous than as in A. viridiilns. Metasoma: 
as in A. viridiilns but terga 1-2 pale amber instead of brown-black. 

Agapostemon angelicus Cockerell 

Agapostemon angelicus Cockerell 1924. Type 9, California Academy of Sciences. 

I have seen the type in San Francisco and it is from Pond Island Bay, 
Angel de la Guarda Island in the Gulf of California, Mexico. As the males 
of A. texanns are rare relative to those of A. angelicus in this region, it 
seems reasonable and expedient to consider this type to be conspecific with 
those males which differ in genitalia and color pattern from those of A. 
texanns. 

Distribution. All males of A. angelicus have been found well within 
the range of the males of A. texanns. Owing to this sympatry and my in- 
ability to distinguish between females of A. texanns and A. angelicus, I have 
here considered the distribution of males and disregarded females. 

Although common only in the arid regions of the southwestern United 
States and northern Mexico, males of A. angelicus have been found as far 
north as west-central North Dakota, as far south as northern Durango 
(Mexico), as far east as central Iowa, and as far west as southwestern 



462 The University Science Bulletin 

California. This species ranges from below sea level in Death Valley, 
California, to 12,000 ft. (3,658 m) on Mount Evans (timberline is at 11,700 
ft.) in Colorado. The altitudinal range of A. angelicas is as astounding as 
the latitudinal range of A. texanus. In Arizona males of A. angelicas have 
been collected from April through November; in Kansas from June through 
October; and in Mexico from June through October. (Map, Fig. 20.) 

Albeit uncommon, A. angelicas does occur in the Great Plains, where 
it is found together with A. texanus in the gallery forests of the tributaries 
of the Missouri and Mississippi Rivers. This habitat is the "Northern 
Floodplain Forest" of Popalas, Salix, and Ulmas (no. 98 in Appendix A). 
While not as diverse as those of A. texanas, its habitats are numerous (39 
of the 116 given by A. W. Kiichler 1964— cf. Appendix A). 

Diagnosis. The male may be distinguished from many other species by 
its toothed hind femora, the apical stylus on its gonostylus, and the lack of 
a low medial ridge on the apical half of its last visible sternum; and from 
A. texanus by the shorter apical stylus with swollen base and enlarged apex 
on its gonostylus, and by the presence of a posterior stripe and absence of 
an anterior stripe on its hind tibia. The female may be distinguished from 
most other species by its bright metallic green to blue metasomal terga, its 
almost hyaline wings, and by the presence of two distinct sizes of 
mesoscutal punctures. 

I have labeled females of A. angelicas as "Agapostemon texanas or A. 
angelicas" because I am unable to separate them from A. texanas. Sand- 
house (1936) claimed that these species differ in the punctation of the 
mesoscutum and in the color of the pubescence. After having examined 
many thousands of females of these species I am forced to conclude that 
the variation within A. texanas nearly encompasses the range of variation 
within A. angelicas, thus invalidating the characters used by Sandhouse. 
It is probably true that the modes of A. angelicas characters differ from those 
of A. texanas but this cannot be demonstrated without positively identified 
females of both species. If one were able to positively identify A. angelicas 
females, perhaps by means of rearing, it might be possible to utilize a 
discriminant function to distinguish between females of the two species. 
Owing to the occurrence of A. texanas males within the range of A. 
angelicas males, I was unable to obtain females which were indisputably 
A. angelicas. Even if one were to find an area where only A. angelicas 
occurred, one could not be sure that differences between these females and 
females of A. texanas were not simply geographic variations. 

Variation. Basing the probable identity of females on the relative 
abundance of males, it would seem that most females of A. angelicas have 
a slightly shinier mesoscutum and whiter pubescence than most of those of 
A. texanas. Like the males, the females of A. angelicas probably average 



Revision of the Bee Genus Agapostemon 463 

slightly smaller than those of A. texanus, but the variation within each 
species is far greater than any difference between them. In both males and 
females of A. angelicas the metallic coloration is the same green as that of 
the sympatric A. texanus. The most noticeably variable character of A. 
angelicits is the amount of melanic pigmentation on the fore coxae of 
males, which range in color from yellow to brown-black, but intra-locality 
variation is far greater than inter-locality variation. 

Description 

male (Figs. 73-74, 164, 181) 

Males as in A. texanus but only about 85% as large with whiter 
pubescence on the mesonotum and metanotum; with dark stripe present 
posteriorly and absent anteriorly on its hind tibia (Fig. 164) ; with apical 
stylus on its gonostylus shorter, swollen basally, and larger apically; with 
medial plate on its gonostylus smaller; and with basal stylus on its 
gonostylus broader and blunt apically (Fig. 181). 

female (Figs. 67-68) 

As in A. texanus but always green and probably smaller with shinier 
mesoscutum and whiter pubescence. 

Agapostemon ascius n. sp. 

The name of this species comes from the Greek sl{ias, or shadow. As 
used in Pliny, ascius literally means without shadow and refers to in- 
habitants of those regions (i.e., the tropics) where the sun is directly over- 
head once or twice each year. 

Agapostemon ascius is known from two females labeled, "5 mi. W. 
Cajamarca/ Tol, COLOMBIA/ 11-20-1965// Figleaf/ gourd/ A. E. & 
M. M./ Michelbacher/ Collectors// 210.1/ 0855-/ 0905." The holotype 
and paratype are at the University of California, Berkeley. 

Diagnosis. A. ascius may be distinguished from most other South Amer- 
ican species by its lack of an acute upturned flange on the anterior margin 
of its mesoscutum, from A. nasutus by the lack of yellow on its posterior 
pronotal lobe, and from the closely related A. leunculus and A. erebus by 
its amber metasoma and 5 (rarely 4?) spatulate teeth on its posterior hind 
tibial spur. 

Description 
female 
General coloration of head and mesosoma bright metallic green, meta- 
soma amber with black at base. Head: (1-4) Labrum, clypeus, interocular 
area and vertex as in A. nasutus. (5) Gena with coarse, parallel and 



464 The University Science Bulletin 

anastomosing carinae. (6) Malar area dark amber to brown, very short. 

(7) Mandible yellowish amber basally, becoming ferruginous on apical %. 

(8) Antenna black above, dark brown below, undersides of first two 
flagellomeres darker than others. Mesosoma: metallic green, never 
with yellow spot on posterior lobe of pronotum as in A. nasutits. (9) 
Pronotum with prominent carina running postero-laterally from lateral 
angle; lateral angle projecting slightly but not markedly as in A. inter- 
medins. (10-12) Mesoscutum, mesoscutellum and metanotum as in A. 
nasutus but punctures slightly coarser and pubescence slightly more dense. 
(13) Mesepisternum very coarsely rugose anteriorly to moderately rugose 
posteriorly; pubescence creamy. (14) Metepisternum as in A. leunculus. 
(15) Propodeum as in A. nasutus but more coarsely sculptured. (16) Wing 
as in A. nasutus. (17) Tegula dark brown with transparent spot postero- 
laterally. (18-20) Fare, middle and hind legs as in A. leunculus. Metasoma: 
(21) Terga amber with anterior half of tergum 1, posterior half of tergum 5 
and lateral margins of 2-4 dark brown to black; pubescence short, sparse and 
inconspicuous. (22) Sterna brown to brown-black with long pale hairs on 
exposed areas. 

Agapostemon atrocaeruleus Friese 

Agapostemon atrocaeruleus Friese 1916. 

I am unable to locate the type of this species but have seen males and 
females labeled "Typus" (which only means they were identified by Friese) 
and they agree with his descriptions. 

Distribution. I have seen specimens from the following localities in 
Costa Rica: Cartago (2 $), San Jose (9 8, 7 2 ), San Mateo (1 8), 
Santiago de Puriscal (12 8,7 5), and Turrialba (7 8 , 12 2). Males 
have been collected from July through November and females from July 
through September. Members of this species have been collected from 
1,950 ft. (594 m) in Turrialba up to 3,963 ft. (1,200 m) in San Jose. 

Diagnosis. The male may be distinguished from males of all Central 
American Agapostemon except A. intermedins by the very acute lateral 
angle and posterior lobe on its pronotum and the conspicuous flange laterally 
on the anterior margin of its mesoscutum; and from males of A. inter- 
medins by the lack of a medial tongue-shaped process, or "button," at the 
base of its 6th metasomal sternum (small bump may be present, but not a 
distinct "button"). The female may be distinguished from females of all 
other species of Agapostemon by the combination of acutely pointed lateral 
pronotal angle and its almost wholly black color (usually with yellow band 
on clypeus and dark metallic blue tints on head and mesosoma). 



Revision of the Bee Genus Agapostemon 465 

Description 

male (Figs. 79-80, 142, 196, 213) 

General coloration of head and mesosoma bright metallic green, 

metasoma banded with black and yellow. Head (Figs. 79-80): pubescence 

golden to amber, becoming creamy on gena. (1) Labrnm as in A. texanus 

but with apical portion shorter and much more acute at apex. (2) Clypeus 

with large scattered punctures on lower (yellow) half, becoming very weakly 

punctate and with weak horizontal carinulae on upper (green) half. (3) 

Interocular area with moderately small and deep contiguous punctures 

below ocelli, becoming shallower and larger (almost rugulose) at level of 

antennal sockets; supraclypeal protuberance with large punctures separated 

by about their own diameters and inclined medially. (4) Vertex with 

moderately small, deep, contiguous punctures anteriorly becoming shallower 

laterally, very weakly rugose between ocelli, and becoming transversely 

carinate posteriorly. (5) Gena with parallel subcontiguous weak carinae 

extending dorso-laterally from antero-ventral margin. (6) Malar area yellow, 

very short. (7) Mandible yellow with apical ! / 3 ferruginous to amber. (8) 

Antenna brown to brown-black above; scape yellow below, pedicel and 

flagellum yellowish to amber below. Mesosoma: pubescence creamy, 

becoming golden to amber on mesonotum and metanotum. (9) Pronotum 

with small, inconspicuous, yellow spot on posterior lobe; lateral angle and 

posterior lobe acutely pointed, lateral angle projecting antero-laterally, 

sharp carina extending postero-ventrally from apex of lateral angle, and 

weak horizontal carinulae postero-ventrally. (10) Mesoscutum with large 

conspicuous flange laterally on anterior margin extending antero-dorsally; 

punctures small (slightly larger than those of A. intermedins), very deep 

and contiguous, becoming rugose antero-laterally. (11) Mesoscutellum with 

medial punctures like those of mesoscutum but separated by about their 

own diameters and becoming larger and much less dense laterally; large 

shiny area with few punctures on each side of mid-line. (12) Metanotum 

coarsely and moderately shallowly rugose. (13) Mesepisternum moderately 

coarsely rugose anteriorly, becoming finely rugose posteriorly with largest 

rugae extending postero-ventrally in arc from anterior margin of metepis- 

ternum. (14) Metepisternttm with coarse, subcontiguous punctures on 

lower Yi-Yi becoming horizontally and irregularly rugose on upper l /i- x /i. 

(15) Propodeum with propodeal carina strong and conspicuous; propodeal 

shield very coarsely and transversely to irregularly rugose (sometimes with 

scattered punctures); dorsal area very coarsely and deeply rugose; lateral 

area with large, deep, subcontiguous punctures and faint rugae (punctures 

and rugae slightly finer anteriorly than posteriorly). (16) Wing hyaline or 

slightly fuliginous, veins dark brown. (17) Tegula transparent amber with 

yellow on proximal margin and submarginally on anterior portion (some- 



466 The University Science Bulletin 

times small amount on posterior margin). (18) Fore leg yellow with dark 
brown basally and posteriorly on coxa; femur with brown apical spot 
dorsally and often with brown streak posteriorly; brown streak posteriorly 
on tibia. (19) Middle leg as fore leg but with coxa dark brown, with brown 
spot at apex of femur larger, and with dark brown streak on posterior of 
tibia extending to dorsal surface. (20) Hind leg (Fig. 142) with coxa bright 
metallic green dorsally and posteriorly, yellow ventrally; trochanter dark 
brown, becoming yellow ventrally; femur yellow with dark brown on 
proximal margin and with dark brown patch at apex; tibia with dark 
brown to brown-black streak antero-dorsally on apical /4-% and with large 
dark brown spot anteriorly on distal margin. Metasoma: (21) Terga 
2-6 dark brown to brown-black on posterior %-% an d yellow on anterior 
Vi'Vi (often partially concealed by overlap); tergum 1 dark brown to brown- 
black with yellow band antero-dorsally (often interrupted laterally); very 
conspicuously punctate with subcontiguous punctures largest on tergum 1, 
becoming progressively smaller on posterior terga. Pubescence golden to 
amber anteriorly on tergum 1, laterally on terga 1-7 and dorsally on terga 
5-7; very short dorsally on terga 1-4, golden to fulvus on yellow bands and 
dark brown to brown-black on dark bands. (22) Sterna yellow, brown 
tinged with green on sternum 1 and with transverse bands on posterior 
margins of sterna 2-5 narrowest on sternum 2, becoming progressively wider 
(especially medially) posteriorly; sternum 6 with slight proximo-medial 
bulge but without distinct "button"; pubescence limited to scattered, moder- 
ately long, fulvus to golden hairs on exposed areas. (23) Genitalia (Figs. 
196, 213) with ventral lobes very short and bearing brush of hairs apically. 

female (Figs. 43-44) 

General coloration of head and mesosoma black or blue-black, pubescence 
black or fuscous. Head (Figs. 43-44): pubescence black or nearly black, 
becoming fuliginous on gena. (1) Labrum as in A. texanus. (2) Clypcus 
with transverse submarginal yellow band distally (sometimes interrupted 
medially or entirely absent); large scattered punctures on lower l / 2 , becoming 
smaller and interspersed with short, weak, horizontal rugulae. (3) Inter- 
ocular area with deep, contiguous, moderately small punctures below ocelli, 
becoming coarsely rugose above antennal sockets, becoming moderately 
coarsely rugose below antennal sockets; supraclypeal protuberance with 
deep, moderately small, scattered punctures and weak horizontal rugae. 
(4) Vertex with moderately small, deep, contiguous punctures anteriorly and 
laterally, becoming smaller and weaker between ocelli; transversely carinate 
posteriorly. (5) Gena with fine, contiguous, parallel carinae extending 
postero-dorsally from antero-ventral margin. (6) Malar area amber; very 
short, almost absent. (7) Mandible yellow, becoming amber to ferruginous 



Revision of the Bee Genus Agapostemon 467 

on basal half. (8) Antenna with scape and pedicel black to brown-black; 
flagellum brown to brown-black above, becoming amber below. Meso- 
soma: pubescence fuscous (white tomentum between posterior lobe of 
pronotum and mesepisternum) becoming black on mesonotum and metano- 
tum. (9) Pronotum with lateral angle and posterior lobe acutely pointed, 
lateral angle projecting antero-laterally; with conspicuous sharp carina ex- 
tending postero-ventrally from apex of lateral angle; with weak irregular 
horizontal rugae postero-laterally. Pubescence above lateral angle and pos- 
terior lobe black. (10) Mesoscutum with conspicuous flange laterally on 
anterior margin extending antero-dorsally; punctures small (slightly larger 
than those of A. intermedins), very deep and contiguous, becoming rugose 
antero-laterally. (11) Mesoscutellum with punctures small, deep and con- 
tiguous medially, becoming widely scattered on shiny lateral area. (12) 
Metanotitm coarsely and shallowly rugose with very deep scattered punc- 
tures. (13) Mesepisternum moderately coarsely rugose anteriorly, becoming 
more finely rugose laterally and with irregular rugae extending antero- 
ventrally from anterior margin of metepisternum. (14) Metepisternum 
horizontally rugulose with few interconnecting rugae. (15) Propodeum 
with propodeal carina very strong and conspicuous; propodeal shield with 
very strong carinae extending dorso-laterally from medial groove; dorsal 
area with very strong, deep rugae extending postero-laterally from anterior 
margin (irregular medially); lateral area with numerous weak horizontal 
carinae anteriorly becoming fewer but larger posteriorly. (16) Wing as 
in $ . (17) Tegula brown to black. (18-20) Fore, middle and hind legs 
brown-black to black; pubescence black dorsally, becoming dark amber 
below. Metasoma: (21) Terga black; punctures separated by about 
their diameters; pubescence very short and appressed dorsally, becoming 
moderately long and erect anteriorly, laterally and posteriorly; pubescence 
black with narrow basal band of sparse white tomentum on terga 2-4. (22) 
Sterna brown-black to black with long black hairs scattered on exposed areas. 

Agapostemon boliviensis n. sp. 

This species is named for the country where it is found. 

The male holotype and paratype as well as the allotype are labeled, 
"BOLIVIA-Beni, Rur-/ renabaque, 175 mts./ 5 October 1956/ (L. Pefia)." 
These types are in the Snow Entomological Museum at the University of 
Kansas. 

Distribution. This species seems to be sympatric with A. mourei. In 
addition to the types, I have seen specimens in the Snow Entomological 
Museum with the following labels: "Cristal Mayu./ Chapare, Cochab/ 
amba Bol. 200 m./ XII-5-49 L E Pena" (2 6 ); same as preceding but col- 
lected the next day (2 8); "BOLIVIA-Dpto. La Paz/ Alta Marani, N. of/ 



468 The University Science Bulletin 

Rurrenbaque, 10 Nov/ 1956 (L. Pena)" (1 S). Two males in the U.S. 
National Museum have the following labels: "Covendo/ Bolivia/ Wm M 
Mann// Aug// Mulford/ BioEpl/ 1921-22" and "nr. mouth/ Rio Mapiri/ 
Sept// Mulford/ Bio Expl. ,, 

Diagnosis. The male may be distinguished from those of most other 
species by the presence of a medial "button" at the base of its last visible 
(6th) sternum (Fig. 137). Agapostemon boliviensis is distinguished from 
the closely related A. inca by its lack of large hooked setae on the postero- 
lateral margin of the antepenultimate visible (4th) sternum, from A. mourei 
and A. inca by the smaller and more rounded medial plate of its gonostylus, 
and from A. intermedins by its enlarged hind femur, tibia and basitarsus 
(Fig. 170). 

The female may be distinguished from most other species by the acute 
lateral angle and posterior lobe of its pronotum and by the dense woolly 
pubescence of its mesoscutum; from those of A. lanosus and A. heterurus by 
the yellow maculations on its clypeus and by its more finely branched meso- 
scutal pubescence; from those of A. mourei by its slightly coarser genal 
striae and its darker pubescence. 

Description 

male (Figs. 137, 170, 177A, 215) 

As in A. mourei but head and mesosoma bright metallic green usually 
with strong, dark copper tints; antenna dark brown to black with scape 
yellow below; tegula dark brown to brown-black; middle leg brown with 
narrow yellow streak anteriorly on trochanter and sometimes on tibia, and 
with yellow anteriorly (and usually posteriorly) on femur; hind leg (Fig. 
170) with ventral ridge of basitarsus more convex; sternum 6 (Fig. 137) 
with proximo-medial projection, or "button," smaller and with each lateral 
flange approximately x / 2 as wide as base (as measured on transverse section 
through widest part of "button"); genitalia (Figs. 177A, 215)." 

female (Figs. 127-128) 

As in A. mourei but with sculpturing slightly coarser and pubescence 
darker; metasoma brown anteriorly, becoming black posteriorly and with 
yellow basally on tergum 3. 

Variation. The color of the head and mesosoma of males ranges from 
a bright metallic blue-green to a dark coppery luster. Furthermore, the 
black markings of the appendages and metasoma are often replaced by pale 
fawn markings. There is no obvious correlation between the color of the 
head and mesosoma and the color of the markings on the appendages and 
metasoma. 



Revision of the Bee Genus Agapostemon 469 

Agapostemon centratus (Vachal) 

Halicttts (Agapostemon) centratus Vachal 1903. Type o* , Museum National LTHistoire 
Naturelle, Paris. 

A metasoma, all that remains of the holotype, is dark brown to black 
with broad bands of white tomentum. I have seen two males labeled "Costa 
Rica" and both have amber metasomas. The genitalia of the type are in- 
distinguishable from those of the males labeled "Costa Rica." The type is 
from Haiti and this species resembles the West Indian A. viridulus group in 
sculpturing and genitalia. Like the similarly labeled specimens of A. 
aenigma, the two males labeled "Costa Rica" are probably from the West 
Indies. 

Diagnosis. This species is unique in the genus inasmuch as the metasoma 
resembles that of females of A. melliventris in being amber to black with 
broad bands of white tomentum basally on the terga. The resemblance is 
so great that Sandhouse misdetermined the two males with amber meta- 
somas as females of A. melliventris. 

Description 

male (Figs. 121-122, 152, 182, 206) 

General coloration of head and mesosoma bright metallic green, meta- 
soma amber to black with conspicuous bands of white tomentum. 
Pubescence of head and mesosoma white to creamy. Head (Figs. 121- 
122): (1) Labrum yellow with transparent margin, evenly rounded apically 
(not acute) ; basal area with conspicuous median depression. (2) Clypeus 
with shallow punctures separated by nearly twice their diameters; pubescence 
sparse, short. (3) Interoadar area shallowly rugose at level of antennal 
sockets, becoming finely rugulose above, sparsely punctate on supraclypeal 
protuberance becoming shallowly rugose laterally; pubescence nearly as 
long as scape medially, becoming shorter laterally and appressed ventro- 
laterally. (4) Vertex with contiguous, fine punctures, becoming rugose be- 
hind ocelli; pubescence like that of interocular area. (5) Genu with weak, 
discontinuous, parallel carinulae running postero-dorsally from antero- 
ventral margin; pubescence dense, nearly 1.5X as long as scape postero- 
medially, becoming shorter anteriorly and laterally. (6) Malar area yellow 
to amber, short, with pubescence white and tomentose. (7) Mandible yellow, 
apical quarter transparent and ferruginous. (8) Antenna. Flagellum dark 
brown above, amber below and on shiny dorsal spot on apex of distal 
fiagellomere; pedicel dark brown above, becoming light brown below; 
scape yellow below, dark brown above. Mesosoma: white pubescence 
moderately dense and evenly distributed (except pronotum). (9) Pronotum 
with lateral angle and posterior lobe rounded, 8-12 parallel carinulae running 
postero-ventrally from below lateral angle; pubescence above lateral angle 



470 The University Science Bulletin 

and posterior lobe like that of mesoscutum, extremely short and fine on 
postero-lateral area. (10) Mesoscutum with fine, contiguous punctures be- 
coming finely rugulose antero-laterally. (11) Mesoscutellum shiny with 
small scattered punctures becoming larger and more numerous on lateral 
margins. (12) Metanotum with very small, sparse punctures medially and 
with weak, short, lateral carinulae directed postero-laterally. (13) Mesepis- 
ternum coarsely but shallowly rugose anteriorly, becoming finely rugose 
posteriorly. (14) Metepisternum finely, regularly and horizontally carinulate. 
(15) Propodeum rounded posteriorly; propodeal carina absent or very weak 
and discontinuous; weak, closely parallel carinulae running ventro-laterally 
from antero-dorsal margin; propodeal shield with few, weak, irregular 
carinulae running laterally and antero-laterally from mid-line. (16) Wing 
transparent and colorless with radius obviously darker than other veins. 
(17) Tegula transparent with metallic green on proximal margin and faint 
yellow transverse band. (18-20) Fore, middle and hind legs (Fig. 152) amber 
to fawn, slightly darker on dorsal surfaces of tibiae of middle and hind 
legs; all coxae metallic green. Pubescence short, creamy to amber. Meta- 
soma: (21) Terga transparent amber with subapical brown band on 
tergum 1; brown band larger, darker and broader on terga 2 and 3 and 
covering exposed areas of terga 4-7. Broad, basal band of dense white 
tomentum on terga 2-5; sparse, moderately long, creamy pubescence on 
anterior surface of tergum 1, ventro-lateral surfaces of terga 1-6, and 
postero-dorsal surfaces of terga 5-7. (22) Sterna transparent, amber, with 
short creamy pubescence on exposed areas directed posteriorly. (23) 
Genitalia (Figs. 182, 206) with weak parallel carinae dorso-laterally on 
gonocoxa. Ventral lobes extremely large, cupped and densely fringed with 
short hairs. 

Agapostemon chapadensis Cockerell 

Agapostemon chapadensis Cockerell 1900. Lectotype o , Carnegie Museum. 

? Agapostemon castanens Schrottky 1902b. Type $ ? Departamento de Zoologia, Sao Paulo? 

Schrottky (1909b) synonymized the male of this species with A. chapadensis. Not having 

seen a type I cannot dispute his judgment. 
? Agapostemon azarae Holmberg 1903. Type $ . 
? Agapostemon argentimts Holmberg 1903. Type $. 
? Agapostemon experiendtis Holmberg 1903. Type $. 

I have seen the syntypes of A. chapadensis and have labeled one male 
lectoholotype, the other male lectoparatype and the female lectoallotype. 
All specimens are from Chapada. I selected a male as the lectoholotype be- 
cause I am unable to distinguish the females of A. semimelleus and A. 
chapadensis, although the female lectoallotypes are easily distinguished. 

Distribution. This species is found in the campos, or tropical grass- 
lands, of southern Brasil, of Paraguay and of northeastern Bolivia. Unlike 



Revision of the Bee Genus Agapostemon 471 

A. semimelleus, it has not been found in the river valleys of the eastern 
slopes of the Andes. 

Diagnosis. The male may be distinguished from those of most other 
South American species by the acute lateral angle on its pronotum and 
lack of a "button" on its last visible sternum (6th). It may be distinguished 
from males of A. semimelleus by the longer distal stylus of its genitalia 
(Figs. 173-174) and by the greater amount of black on the posterior of its 
hind femur (Figs. 171-172). 

The female cannot be distinguished from that of A. semimelleus (cf. 
A. semimelleus, Variation) but it may be distinguished from those 
of other South American Agapostemon by the acute upturned flange on 
its anterior mesoscutal margin and lack of basal yellow bands on its meta- 
somal terga. 

Description 
male (Figs. 172-173, 219) 

General color of head and mesosoma bright metallic green, metasoma 
banded with black and yellow. Like male of A. semimelleus but for the 
following features. Mesosoma: (18) Fore leg yellow with brown (some- 
times with slight metallic tints) posteriorly on trochanter, femur and tibia. 
(19) Middle leg yellow with brown on coxa, basitarsus, posterior of 
trochanter and tibia, and posterior of apical x / z of femur. (20) Hind leg 
(Fig. 172) yellow with upper x / 2 of coxa metallic, with brown on trochanter, 
on posterior and apex of femur, on posterior and antero-dorsal regions of 
tibia, and on all but antero-ventral region of basitarsus. Metasoma: 
(22) Sterna 2-5 brown with amber posterior margins and sometimes with 
yellow on anterior margins, sternum 1 amber with metallic tints basally, 
and sternum 6 amber becoming brown on anterior margin. (23) Genitalia 
(Figs. 173, 219) ventral lobes very short with apical brush of long stiff 
hairs. Distal stylus about twice as long as width of medial plate. 

FEMALE 

Like that of A. semimelleus (cf. A. semimelleus, Variation). 
Agapostemon cockerelli Crawford 

Agapostemon cockerelli Crawford 1901. Type $ , U.S. National Museum. 
Agapostemon martini Cockerell 1927. Males only (misidentifieil). 

I have examined the types of Agapostemon cockerelli Crawford and 
Agapostemon jemoratus Crawford (1901) in Washington, D.C. and, con- 
trary to the opinions of Sandhouse (1936) and Michener (1951), find them 
specifically distinct. The males (paratypes) of Agapostemon martini Cock- 
erell (1927) were misdetermined and belong in A. cockerelli. Nomia cillaba 
Cameron (1902), listed by Michener {In Muesebeck, et al., 1951) as a 



472 The University Science Bulletin 

possible synonym of Agapostemon coc\erelli, does not belong in the genus 
Agapostemon (cf. Excluded Species). 

Distribution (Fig. 2). Agapostemon cockerelli occurs as far north as 
Pat Creek, Park Co., Wyoming; as far south as Petlalcingo, Puebla, Mexico; 
as far east as Tehuacan, Mexico; and as far west as Wickenburg, Maricopa 
Co., Arizona. Most abundant at elevations of 4,500-7,500 ft. (1,372-2.286 m), 
this species has been collected as high as 8,000 ft. (2,438 m) on Pine Top 
Mountain, Culberson Co., Texas, and as low as 2,100 ft. (640 m) in Big 
Bend National Park, Brewster Co., Texas. In Mexico it is apparently re- 
stricted to the central plateau, where it has been collected as high as 7,300 
ft. (2,225 m) 20 miles north of Zacatecas, Zacatecas, and as low as 1,850 ft. 
(564 m) in Boquillas del Carmen, Coahuila. In the United States females 
of A. cockerelli have been collected from March through October and males 
from April through November. In Mexico females have been collected from 
June through October and males from April through December. 

Diagnosis. The male may be distinguished from males of many species 
by its toothed and conspicuously inflated hind femora, the lack of an 
apical stylus on its gonostylus, or by its nearly hyaline wings; from males 
of A. radiatus by the basal ridge on its basitarsus; and from males of A. 
femoratus by the much smaller basal ridge on its hind basitarsus (Figs. 
162-163). The female may be distinguished from females of many species 
by its metallic green metasomal terga, its coarsely punctate or rugose meso- 
scutum without punctures of two distinct sizes, and its nearly hyaline 
wings; and from females of A. femoratus by its largely punctate (rugose 
only anteriorly and laterally) mesoscutum. The female cannot be mor- 
phologically differentiated from females of A. radiatus' but may be dis- 
tinguished by its southwestern distribution. 

Description 

male (Figs. 107-108, 162) 

General color of head and mesosoma bright green to blue-green, meta- 
soma with black and yellow bands. Head (Figs. 107-108): pubescence 
white, becoming pale yellowish on vertex and sometimes on interocular 
area; more dense than in A. radiatus. (1) Labrum as in A. texanus. (2-5) 
Clypeus, interocular area, vertex and gena as in A. radiatus but with slightly 
coarser and shallower sculpturing. (6-7) Malar area and mandible as in 
A. radiatus. (8) Antenna as in A. radiatus but with underside of flagellum 
pale amber to yellowish and with upper side of apical half of apical flagello- 
mere pale amber to yellow. Mesosoma: pubescence white, commonly 
becoming yellowish on mesonotum and metanotum. (9) Pronotum with 
lateral angle and posterior lobe slightly more angular than in A. radiatus 
and with sculpturing weaker than in A. radiatus. (10-15) Mesoscutum, 



Revision of the Bee Genus Agapostemon 473 

mesoscutelhim, metanotum , mesepisternum, metept sternum and propodeum 
as in A. radiatus but with sculpturing shallower and slightly finer. (16-17) 
Wing and tegida as in A. radiatus but slightly paler. (18-19) Fore and 
middle legs as in A. radiatus but with brown maculations reduced (may be 
absent on fore leg). (20) Hind leg (Fig. 162) as in A. radiatus but with 
trochanter yellow with femur and tibia more swollen and with small basal 
ridge and apical groove on basitarsus. Metasoma: (21-22) Terga and 
sterna as in A. radiatus but with dark brown or black areas reduced, and 
with stronger metallic tints postero-laterally on terga 3-4. (23) Genitalia 
indistinguishable from those of A. jemoratus. 

female (Figs. 59-60) 

General coloration of head, mesosoma and metasoma bright metallic 
green to blue-green. Head (Figs. 59-60) : pubescence white, commonly 
with faint yellow tints on vertex. (1) Labrum as in A. texanus. (2-4) 
Clypeus, interocular area and vertex as in A. radiatus but with sculpturing 
slightly shallower and slightly finer. (5) Gena as in A. radiatus but with 
genal carinae usually slightly coarser. (6-7) Malar area and mandible as in 
A. radiatus. (8) Antenna as in A. radiatus but often paler on underside of 
flagellum and with yellowish amber at base and apex of scape. Meso- 
soma: pubescence white, becoming slightly yellowish to fuliginous on 
mesonotum and metanotum. (9) Pronotum as in A. radiatus but with 
sculpturing slightly weaker. (10-13) Mesoscutum, mesoscutelhim, metano- 
tum, mesepisternum as in A. radiatus but with sculpturing slightly coarser. 
(14) Metepisternum as in A. radiatus but irregularly anastomosing carinae 
with greater tendency to rugose condition. (15) Propodeum as in A. radiatus 
but with slightly finer rugae on propodeal shield; dorsal surface with much 
shallower sculpturing and more commonly rugose or with more frequently 
anastomosing carinae. (16) Wing as in A. radiatus but slightly paler. (17) 
Tegula as in A. radiatus £ but slightly paler. (18-20) Fore, middle and 
hind legs as in A. radiatus but with pubescence paler; with yellow markings 
on fore leg and middle leg; and usually with more extensive yellow mark- 
ings anteriorly on fore tibia. Metasoma: (21) Terga as in A. radiatus 
but bands of white tomentum on terga 2-4 slightly broader and more dense. 
(22) Sterna as in A. radiatus but rarely with weak metallic tints on 
sternum 4. 

Agapostemon coloradinus (Vachal) 

Agapostemon coloradensis Crawford 1901. Lectotype* 9 , U.S. National Museum. 
Halictus (Agapostemon) coloradinus Vachal 1903. 

When Vachal placed Agapostemon in the genus Halictus, Agapostemon 

* The specimen is labeled "Lectotype." Although I have not been able to find a lectotype 
designation in the literature, I accept this specimen as a lectotype and formalize it by so 
designating it here. 



474 The University Science Bulletin 

coloradensis Crawford became a junior secondary homonym of Augochlora 
coloradensis Titus. Vachal renamed the former, Halictus (Agapostemon) 
coloradinus. Few workers agreed with Vachal's placement of Agapostemon 
as a subgenus of Halictus, but the International Code of Zoological Nomen- 
clature (anonymous, 1964) states (Article 59c) that secondary homonyms 
rejected before 1961 cannot be revived, contrary opinions (Cockerell, 1937a 
and Michener, In Muesebeck, et al., 1951) notwithstanding. 

Sandhouse (1936) included Agapostemon tyleri Cockerell and Aga- 
postemon martini Cockerell in Agapostemon coloradinus (Vachal), but hav- 
ing examined the types I find her decisions unjustified (cf. A. tyleri). 

Distribution (Fig. 3). This is by far the rarest of the North American 
species and its range is more restricted than that of any other North Ameri- 
can species. I have seen fewer than 300 specimens, from the high, western 
plains of South Dakota, Nebraska, Kansas, Oklahoma and Texas and from 
both the plains and mountains of New Mexico, Colorado, Wyoming and 
central Utah. Females have been collected from May through October and 
males from August through October. Specimens have been collected as low 
as 2,060 ft. (628 m) in Rush Co., Kansas and as high as 10,600 ft. (3,231 m) 
at Echo Lake, Mt. Evans, Colorado. 

Diagnosis. The male may be distinguished from all other North Ameri- 
can species except A. tyleri by the lack of yellow maculations on its dark 
brown to black metasomal sterna; from A. tyleri by its larger size, by the 
brown streak on the central % of the length of the posterior surface of its 
hind tibia (Figs. 159-160), and by its more slender penis valve which has a 
conspicuous row of hairs on its dorsal crest (Figs. 226-227). The female 
may be distinguished from many other species occurring north of Mexico 
by its black metasoma and the lack of yellow on its clypeus; from A. 
virescens by its finer and more numerous genal carinulae (2-3 per 0.25 mm 
in A. virescens and 5-6 per 0.25 mm in A. coloradinus) ; and from A. tyleri 
by its brown wings, paler tegulae and less shiny supraclypeal protuberance. 
This species is very close phenetically to A. tyleri. 

Description 

male (Figs. 99-100, 160, 185, 199, 226) 

General coloration of head and mesosoma bright metallic blue to yreenish 
blue, metasoma with black and pale yellow to creamy bands. Head (Figs. 
99-100): pubescence snowy white. (1) Labrum as in A. texanus but more 
rounded at apex. (2) Clypeus as in A. virescens but punctures shallower 
and pubescence more dense. (3) Interocular area with coarse contiguous 
punctures (some specimens finely rugose around antennal sockets) ; supra- 
clypeal protuberance as in A. virescens. (4) Vertex as in A. virescens. (5) 
Gena as in A. virescens but with carinae much finer. (6) Malar area yellow, 



Revision of the Bee Genus Agapostemon 475 

never amber; short. (7) Mandible as in A. virescens. (8) Antenna dark 
brown to brown-black above with apex of apical flagellomere amber; 
flagellum amber below and underside of scape and pedicel yellow (scape 
of some specimens from Boulder, Colorado, area black with yellow apically 
on underside). Mesosoma: pubescence snowy white. (9) Pronotum as 
in A. virescens but with weaker sculpturing. (10) Mesoscutum as in A. 
virescens but with finer punctures and with anterior margin punctate. 
(11-13) Mesoscutellum , metanotum and mesepistemum as in A. virescens 
but with finer sculpturing. (14) Metepisternum rugose or with irregularly 
anastomosing carinae. (15) Propodentn with propodeal carina weak, often 
broadly interrupted laterally; propodeal shield finely and weakly rugulose; 
finely rugose dorsally and postero-laterally becoming punctate antero- 
laterally. (16) Wing as in A. virescens but paler and with darker radial 
vein. (17) Tegnla as in A. virescens but slightly paler and transparent. 
(18-19) Fore and middle legs as in A. virescens but with brown-black mark- 
ings of tibiae and femora more extensive and sometimes with yellow apically 
on ventral surface of trochanter. (20) Hind leg (Fig. 160) as in A. virescens 
but tibia with brown streak on apical Y 2 of posterior surface, femur with 
brown streak entire length of posterior surface, trochanter usually with 
yellow spot apically on antero-ventral surface, and tooth on femur directed 
distally. Metasoma: (21) Terga brown-black with creamy bands on 
basal halves of terga 2-5 and centrally on tergum 1 (commonly interrupted 
medially on tergum 1); pubescence short dorsally on terga 1-4, black on 
dark bands and white on creamy bands; moderately long and white else- 
where. (22) Sterna brown to brown-black, usually with faint carina medially 
on apical l / 2 of sternum 6; moderately long white hairs scattered on exposed 
areas. (23) Genitalia (Figs. 185, 199, 226) as in A. virescens but with 
smaller fold on medial plate of gonostylus and with conspicuous hairs on 
prominent dorsal crest of penis valve. 

female (Figs. 49-50) 

General coloration of head and mesosoma bright metallic blue to blue- 
green, metasoma black with white hair bands. Head (Figs. 49-50): (1) 
Labrum as in A. texanus but with distal keel slightly broader. (2-4) Clypeus, 
interocular area and vertex as in A. virescens but with sculpturing slightly 
finer. (5) Gena as in A. virescens but with much finer and more numerous 
(5-6 per 0.25 mm) carinulae. (6) Malar area dark ferruginous to brown- 
black; very short, almost absent. (7) Mandible very dark amber with apical 
l / 3 darker ferruginous and base almost black (with very faint metallic tints 
on some specimens). (8) Antenna brown-black, lower half of flagellum 
usually slightly paler brown than upper l / 2 . Mesosoma: snowy white 
pubescence sometimes very slightly fuliginous on mesonotum and meta- 



476 The University Science Bulletin 

notum. (9-15) Pronotum, mesoscutum, mesoscutellum , metanotum , mese- 
pisternum, metepisternum and propodeum as in A. virescens but with 
sculpturing slightly shallower and finer. (16) Wing brown in contrast to 
that of male; unlike that of A. splendent, not markedly darker on distal 
borders; radius brown-black, other veins and pterostigma dark brown. 
(17) Tegida as in A. virescens. (18-20) Fore, middle and hind legs as in 
A. virescens but with pubescence basally on antero-dorsal surface of hind 
leg slightly paler (usually amber but may be dark brown). Metasoma: 
(21-22) Terga and sterna as in A. virescens. 

Agapostemon columbi n. sp. 

This species is named after Christopher Columbus who first landed in 
the New World on San Salvador Island (=Watling I.). 

I have seen one male and one female of A. columbi. The male is labeled, 
"Watlings [sic] Is./Bahama Islands/W. W. Worthington/III. 13.1909/7 
Cam. Mus./Acc. 3960." The female is labeled, "Wading Is./Bahama 
Islands/W. W. Worthington/III.20.1909//Carn. Mus./Acc. 3960." The 
male holotype and the allotype are both at the Carnegie Museum, Pittsburgh, 
Pennsylvania. 

Diagnosis. The female may be distinguished from A. viridulus, A. 
obscurants, A. hispaniolicus, A. swainsonae , A. alayoi and A. hohliellus by 
its metallic metasomal terga; from A. sapphirinus, A. cyaneus and A. 
ochromops by its lack of milky yellowish eyes and pterostigma; from A. 
viequesensis by the metallic green tints on the base of its mandible; from 
A. poeyi and A. insularis by its lack of metallic green tints on metasomal 
sterna 3 and 4; and from A. aenigma by the golden (rather than white) 
pubescence on the undersides of its hind femur and tibia. 

The male may be distinguished from A. viridulus, A. obscurants, A. 
hispaniolicus, A. swainsonae and A. cubensis by its lack of an elongate 
clypeal area; from A. \ohliellus and A. centratus by the tooth on its hind 
femur; from A. sapphirinus, A. cyaneus and A. ochromops by its lack of 
milky to yellowish eyes and pterostigma; from A. insularis, A. jamaicensis 
and A. poeyi by its lack of metallic green or blue medially on metasomal 
tergum 3; and from A. viequesensis by its conspicuous metallic green tints 
medially on metasomal terga 4-5. Both males and females of A. columbi 
are slightly larger than those of other members of the A. viequesensis 
species group. 

Description 

MALE 

General coloration of head and mesosoma bright metallic blue-green, 
metasoma dark brown with narrow yellow bands. Head: as in A. vie- 
quesensis but with pubescence amber on interocular area and vertex. 



Revision of the Bee Genus Agapostemon 477 

Mesosoma: pubescence on mesonotum and metanotum pale amber, else- 
where pale yellowish to creamy. (9) Pronotum as in A. vie que sen sis but 
with fewer and coarser horizontal carinae on postero-lateral area, and with 
lateral angle and posterior lobe slightly more angular. (10-14) Mesoscutum, 
mesoscutelhim , metanotum , mesepisternum and metepisternum as in A. vie- 
quesensis but with coarser sculpturing as in A. poeyi (but not as coarse). (15) 
Propodeum with sculpturing much coarser than in A. viequesensis and with 
prominent carinae extending postero-laterally from posterior margin of 
metanotum. (16) Wing as in A. viequesensis but slightly darker and with 
darker veins and pterostigma. (17) Tegula as in A. viequesensis but with 
transparent amber portion slightly darker. (18-20) Fore, middle and hind 
legs as in A. viequesensis. Metasoma: pubescence as in A. viequesensis. 
(21) Terga. Tergum 1 dark amber on anterior l / 2 , dark brown on posterior 
dorsal area, and with yellow band centrally, broadly interrupted medially. 
Terga 2-3 dark brown with yellow band on anterior l / 3 . Terga 4-7 like 
2-3 but with conspicuous metallic tints on dark brown areas. (22) Sterna 
as in A. viequesensis. (23) Genitalia as in A. viequesensis but with apex of 
distal stylus on gonostylus slightly enlarged, but not so much as in A. poeyi. 

FEMALE 

General coloration of head, mesosoma and metasoma bright metallic 
blue-green to blue (metasoma with purple-brown tints). Head: pubes- 
cence white, only very faintly fuliginous on vertex. (1) Labrum as in A. 
hispaniolicus (cf. A. viequesensis). (2-5) Clypeus, interocular area, vertex 
and gena as in A. viequesensis but with sculpturing much coarser, although 
not as coarse as in A. poeyi. (6-8) Malar area, mandible and antenna as in 
A. viequesensis but with metallic spot at base of mandible. Mesosoma: 
as in A. viequesensis but pubescence paler on mesonotum and metanotum. 
(9) Pronotum with lateral angle and posterior lobe slightly more angular 
than in A. viequesensis, although not so much as in A. poeyi; and with 
postero-lateral horizontal carinae fewer in number and coarser than in A. 
viequesensis although not as few and as coarse as in A. poeyi. (10) Meso- 
scutum with punctures slightly coarser than in A. viequesensis, but not 
nearly as coarsely sculptured as in A. poeyi. (11) Mesoscutelhim as in A. 
viequesensis but less shiny and with punctures deeper and denser. (12-15) 
Metanotum, mesepisternum, metepisternum and propodeum as in A. poeyi 
but with finer sculpturing, although not as fine as in A. viequesensis. (16) 
Wing as in $ . (17) Tegula as in A. viequesensis. (18-20) Fore, middle and 
hind legs as in A. viequesensis but with pubescence amber, becoming brown- 
black antero-dorsally on tibia and basitarsus (not as dark on latter) of hind 
leg. Metasoma: as in A. viequesensis but with metallic areas bluer and 
nonmetallic areas darker. 



478 The University Science Bulletin 

Agapostemon cubensis n. sp. 

This species is named for the island of Cuba where it is found. 

I have seen a single male labeled, "Cuba 5.K./S. of Pinar Rio,/ Sept. 
12-23 '13." The type is in the American Museum of Natural History, New 
York. 

Diagnosis. The male of this species may be distinguished from A. 
kohliellus and A. centratus by its toothed hind femora; from A. viridulus, 
A. obscuratus and A. hispaniolicus by its pale amber and brown banded 
basal metasomal terga; from A. swainsonae by its small and narrowly 
grooved basitarsal ridge; and from males of A. poeyi, A. vie que sen sis, A. 
cyaneus, A. sapphirinus, A. ochromops, A. columbi, A. insularis and A. 
jamaicensis by its elongate clypeal area. 

Description 

male (Figs. 119-120, 149) 

General coloration of head and mesosoma shiny, metallic blue and 
green, metasoma pale amber and brown. Head (Figs. 119-120): pubes- 
cence white. (1) Labrum as in A. viridulus but white and darkened only 
centrally. (2) Clypeus metallic blue above, white below; elongate as in A. 
viridulus; shiny and sparingly punctate medially, becoming more densely 
punctate laterally; pubescence sparse. (3) Interocular area metallic blue; 
shallow contiguous punctures below median ocellus, shallowly and finely 
rugose just above and at level of antennal sockets, becoming shiny with 
scattered punctures below antennal sockets, supraclypeal protuberance shiny 
and sparsely punctate and with weak transverse carinulae; pubescence be- 
tween antennal sockets about % as long as scape, becoming shorter laterally, 
dorsally and ventrally; tomentose ventro-laterally, near malar area. (4) 
Vertex metallic blue, changing to blue-green posteriorly; punctures very fine 
(finer than in A. viridulus), shallow anteriorly and laterally, becoming 
rugose posteriorly; area between ocelli sparsely and shallowly punctate; 
pubescence white, not fuliginous as in A. viridulus. (5) Gena metallic blue- 
green in contrast to metallic blue of clypeus, interocular area and vertex; 
rugae as in A. viridulus but weaker; pubescence as long as scape postero- 
medially, becoming shorter laterally, anteriorly and dorsally. (6-8) Malar 
area, mandible and antenna as in A. viridulus but with white replacing 
yellow. Mesosoma: pubescence as in A. viridulus but white on meso- 
notum and metanotum. (9) Pronotum as in A. viridulus but blue dorso- 
medially, becoming blue-green ventro-laterally. (10) Mesoscutum metallic 
blue with very fine punctures, slightly shallower and less dense than in 
A. viridulus and becoming weakly rugulose anteriorly. (11) Mesoscutellum 
metallic blue-green, punctures as in A. viridulus but slightly shallower. 
(12) Metanotum metallic blue-green, with sculpturing as in A. viridulus. 



Revision of the Bee Genus Agapostemon 479 

(13) Mesepisternum metallic blue-green antero-dorsally becoming green 
posteriorly and ventrally, with sculpturing as in A. viridulus but slightly 
coarser. (14) Metepisternum metallic blue-green dorsally, becoming green 
ventrally, with sculpturing as in A. viridulus but slightly coarser. (15) 
Propodeum metallic green, with sculpturing as in A. viridulus. (16) Wing 
as in A. viridulus but with veins not as dark. (17) Tegula as in A. viridulus 
but transparent portion amber and metallic tints blue. (18-19) Fore and 
middle legs as in A. viridulus but with pale amber replacing dark brown. 
(20) Hind leg (Fig. 149) with coxa and trochanter as in A. viridulus; femur 
amber with white antero-ventrally, on basal % antero-dorsally, and on 
postero-ventral margin between tooth and distal margin; tibia pale amber 
ventrally becoming pale brown dorsally; tarsus amber. Femur less swollen 
than in A. viridulus, basitarsus with narrow basal ridge much smaller than 
in A. viridulus, apical groove much narrower and shallower than in A. 
viridulus. Pubescence as in A. viridulus. Metasoma: (21) Terga. Ter- 
gum 1 pale amber with narrow band of pale brown anteriorly on dorsal 
surface; tergum 2 pale amber with central brown band darker and broader 
than on preceding tergum 1; tergum 3 with broad dark brown band cen- 
trally, pale amber basally, and brown apically; terga 4-7 brown-black basally, 
becoming brown apically. Moderate to long white hairs scattered on ventro- 
lateral margins, anteriorly on tergum 1, and posteriorly on terga 5-7; short, 
dense, simple, prostrate hairs on terga 1-5 brown and inconspicuous; thin, 
white inconspicuous tomentum basally on terga 1-6 (usually hidden by 
preceding terga). (22) Sterna as in A. viridulus but amber with subapical, 
pale brown band narrow on sternum 3, becoming broader and darker on 
sterna 4-6. (23) Genitalia (examined dry) as in A. viridulus but with about 
l / 2 as many grooves on gonocoxa. 

Agapostemon cyaneus n. sp. 

The name of this species refers to the dark blue color of the female. 

I have seen ten specimens (7 3,3 2 ) of A. cyaneus. The female holo- 
type, the allotype and seven paratypes are labeled, "Bahamas/Crooked Isl./ 
1-2 Mar//Utowana/Exp. 1934," and one male paratype is labeled, "Ba- 
hamas/Crooked Isl./20-II//Greenway/coll." All of the types are in the 
Museum of Comparative Zoology, Harvard University, Cambridge, Mas- 
sachusetts. 

Diagnosis. Both males and females may be distinguished from all other 
West Indian species, except A. sapphirinus and A. ochromops, by their 
milky to yellowish eyes and pterostigma. The male of A. cyaneus may be 
distinguished from A. sapphirinus and A. ochromops by its pedicel which 
is yellow to amber below (not dark brown). The female of A. cyaneus may 
be distinguished from A. ochromops by its very dark blue color, and from 
A. sapphirinus by its finely carinate gena. 



480 The University Science Bulletin 

Description 

male 

General coloration of head and mesosoma dark metallic green to blue, 
metasoma black with pale yellow bands. Head: darker than in A. vie- 
quesensis and with paler eyes; pubescence as in A. viequesensis. (1-7) La- 
brum, clypeus, interocular area, vertex, gena, malar area and mandible as in 
A. viequesensis. (8) Antenna as in A. viequesensis but with black on upper 
surface replaced by yellow on distal l / 2 of apical flagellomere and basal Vr 1 /* 
of scape, with underside of flagellum pale amber to yellow, and with un- 
derside of scape dark brown to black. Mesosoma: pubescence as in A. 
viequesensis. (9-15) Pronotum, mesoscutum, mesoscutellum , metanotum, 
mesepisternum, metepisternum and propodeum as in A. viequesensis but 
with very slightly coarser sculpturing. (16) Wing as in A. viequesensis but 
with pale yellow pigment on pterostigma and underside of larger veins. 
(17) Tegula as in A. viequesensis but with amber replaced by dark brown. 
(18-20) Fore, middle and hind legs as in A. viequesensis but with brown- 
black replacing amber, with yellow areas paler, with brown-black postero- 
dorsally on apex and base of middle femur (and sometimes fore femur), 
and with average size of basal ridge of basitarsus slightly greater. Meta- 
soma: pubescence as in A. viequesensis. (21) Terga brown-black with 
narrow, pale yellow bands basally on terga 2-7 (those on posterior terga 
usually hidden by overlap), and centrally on tergum 1 (often interrupted 
medially); apex of pygidium nearly as acute as in A. ochromops (Fig. 225). 

(22) Sterna as in A. viequesensis but brown-black to, amber, and sometimes 
with metallic spot medially on transverse submarginal ridge of sternum 4. 

(23) Genitalia as in A. poeyi but distal stylus on gonostylus not as swollen 
at apex (cf. A. viequesensis). 

FEMALE 

General coloration of head, mesosoma and metasoma very dark metallic 
bluish purple. Head: as in A. viequesensis but with eyes whitish, pubes- 
cence wholly white, mandibles brown-black with metallic tints at base and 
ferruginous apically. Mesosoma: as in A. viequesensis but with pubes- 
cence of mesonotum and metanotum white and fuliginous dorsally on mid- 
dle tibia, hind tibia and apex of hind femur; with wings as in S ; and with 
tegula brown-black with metallic tints antero-basally. Metasoma: as in 
A. viequesensis but with basal bands of white tomentum on terga broader 
and denser, and with sterna very dark brown with faint metallic tints. 

Agapostemon erebus n. sp. 

This species is named after the Greek god Erebus, personification of 
darkness and brother of Night. 



Revision of the Bee Genus Agapostemon 481 

The female holotype is in the California Academy of Sciences, San 
Francisco. It is labeled, "El Salvador/Mt. San Salvador/ VII-8-1963//M. E. 
Irwin &/D. Q. Cavagnaro/Collectors." A second female is in the University 
of Kansas collection labeled, "Guatemala 6.3 mi./NE. Escuintla/30 July 
1966/U. Kansas Mex. Exped." 

Diagnosis. A. erebus may be distinguished from the similarly colored 
females of A. atrocaeruleus by its rounded lateral pronotal angle and its 
coarse genal carinae. It may be distinguished from those females of A. 
nasutus with dark blue head and thorax by the prominent carina running 
postero-ventrally from the lateral angle of its pronotum, by the lack of 
yellow on the posterior lobe of its pronotum, and by its coarse genal carinae. 
It seems most closely related to A. leunculus but differs from females of this 
species by its blue-black head and thorax and shiny lateral area on its 
mesoscutellum. 

Description 

FEMALE 

General coloration of head and mesosoma very dark blue to black and 
only slightly metallic, metasoma black with narrow bands of white pubes- 
cence dorsally. Head: (1-4) Labrum, clypeus, interocular area and vertex 
as in A. nasutus but with pubescence predominantly black. (5) Gena with 
coarse, parallel and anastomosing carinae and whitish pubescence as in A. 
leunculus. (6) Malar area black, short. (7) Mandible yellow basally, be- 
coming ferruginous on apical %• (8) Antenna as in A. leunculus. Meso- 
soma: (9) Pronotum as in A. nasutus but dark blue and with prominent 
carina running postero-ventrally from lateral angle. (10) Mesoscutum dark 
purple-blue marginally, dull green-black centrally; sculpturing as in A. 
nasutus but slightly finer; pubescence black. (11) Mesoscutellum dark 
purple-black; punctation as in A. nasutus but with lateral shiny area more 
shiny and conspicuous; pubescence black. (12) Metanotum dark purple- 
black, rugulose; pubescence black. (13) Mesepisternum purple-black; sculp- 
turing as in A. leunculus; pubescence dark fuliginous on dorsal l / } , becoming 
white below. (14-15) Metepisternum and propodeum dark blue, sculpturing 
as in A. nasutus but slightly coarser; pubescence white, sparse except around 
and posterior to propodeal spiracle. (16) Wing transparent, not darkened 
apically, radius scarcely darker than other veins. (17) Tegula black, dark 
transparent brown postero-laterally. (18) Fore leg brown-black with small 
yellow spot at base of tibia; pubescence of coxa and trochanter white, re- 
mainder dark brown to black. (19-20) Middle and hind legs brown-black; 
pubescence similar in color to that of fore leg, but scopal hairs creamy. 
Metasoma: (21-22) Terga and sterna as in A. leunculus (on the holotype 
the normally hidden, pale brown basal areas of terga are exposed). 



482 The University Science Bulletin 

Agapostemon femoratus Crawford 

Agapostemon femoratus Crawford 1901, Type $ , U.S. National Museum. 
Agapostemon calijornicus Crawford 1901. Females only (misidentified). 

Sandhouse (1936) and Michener (1951) both considered A. femoratus 
synonymous with A. cocl^erelli Crawford (1901), but after examining the 
types of both species in Washington, D.C., I concluded that they are spe- 
cifically distinct. The females (paratypes) of Agapostemon calijornicus 
Crawford (1901) were misdetermined and belong in A. femoratus. 

Distribution. Agapostemon femoratus occurs as far north as Chilcotin, 
British Columbia, Canada; as far south as San Vicente, Baja California, 
Mexico; as far west as the coast of California; and as far east as Williston, 
North Dakota. In the northern part of the range (Washington), females 
have been collected from April through October and males from June 
through October. In California south of Los Angeles, females have been 
collected from February through November, and males in March and May 
through November. This species is equally abundant at high and low eleva- 
tions. A. femoratus has been collected as high as 10,150 ft. (3,094 m) at 
Blanco's Corral, Mono Co., California; 10,000 ft. (3,048 m) at Blue Ridge 
Fire Lookout, Fremont Co., Wyoming; 9,200 ft. (2,804 m) at Warner 
Ranger Station, Grand Co., Utah; and 8,600 ft. (2,541 m) on Hart Mountain, 
Lake Co., Oregon. (See map. Fig. 4.) 

Diagnosis. The male may be distinguished from many other species by 
its conspicuously inflated and toothed hind femora, the large and conspicuous 
apical groove on its hind basitarsus, or its nearly hyaline wings; from A. 
radiatus by the large basal ridge on its basitarsus; and from A. cocherelli 
by its much larger basal ridge on its hind basitarsus and much more in- 
flated hind leg (Figs. 162-163). The female may be distinguished from 
other species by its metallic green metasomal terga or its coarsely rugose 
mesoscutum. 

Description 

male (Figs. 101-104, 163, 194, 202) 

General color of head and mesosoma bright metallic green, metasoma 
with black and yellow bands. Head (Figs. 101-104): pubescence white, 
usually yellowish on vertex and often pale yellow on interocular area, 
denser than in A. radiatus. (1) Labrum as in A. texanus. (2-7) Clypeus, 
interocular area, vertex, gena, malar area and mandible as in A. radiatus. 
(8) Antenna as in A. radiatus but with underside of flagellum pale amber 
to yellowish and with upper side of distal |/ 2 of apical flagellomere pale 
amber to yellow. Mesosoma: pubescence white, commonly becoming 
yellowish on mesonotum and metanotum. (9) Pronotum as in A. radiatus 
but with lateral angle and posterior lobe slightly more angular, and sculp- 



Revision of the Bee Genus Agapostemon 483 

turing weaker. (10-15) Mesoscutum, mesoscutellum, metanotum, mesepi- 
sternum, metepistemum and propodeum as in A. radiatus, but with sculp- 
turing slightly shallower. (16-17) Wing and legula as in A. radiatus but 
paler. (18) Fore leg as in A. radiatus but lacking brown stripe on tibia. 

(19) Middle leg as in A. radiatus but with brown tibial stripe very reduced. 

(20) Hind leg (Fig. 163) as in A. radiatus but with trochanter yellow; brown 
stripe on tibia reduced; femur, tibia and basitarsus swollen (femur sub- 
globose) ; and basitarsus with very large, slightly sinuate basal ridge and 
large broad apical groove. Metasoma: (21-22) Terga and sterna as in 
A. radiatus but with dark brown or black areas reduced, and with stronger 
metallic tints postero-laterally on terga 3-4. (23) Genitalia (Figs. 194, 202) 
with gonocoxite partially fused with gonobase; gonostylus with large medial 
plate but lacking apical and basal styli ; ventral lobe of gonocoxite moderately 
large and with distal fringe of hairs. 

female (Figs. 57-58) 
General coloration of head, mesosoma and metasoma bright metallic 
blue-green to blue. Head (Figs. 57-58): pubescence white, commonly 
very pale yellow on vertex. (1) Labrum as in A. texanus. (2-4) Clypeus, 
interocular area and vertex as in A. radiatus but with sculpturing slightly 
deeper and more coarse. (5) Gena as in A. radiatus. (6) Malar area dark 
brown to brown-black; short. (7) Mandible as in A. radiatus but with 
pigment creamy and less opaque. (8) Antenna as in A. radiatus but with 
flagellum slightly paler on underside. Mesosoma: pubescence white, be- 
coming very faintly yellowish on some specimens. (9) Pronotum as in 
A. radiatus but with slightly finer sculpturing. (10) Mesoscutum very 
coarsely and deeply rugose, becoming coarsely and contiguously punctate 
postero-medially. (11-15) Mesoscutellum , metanotum, mesepisternum, me- 
tepistemum and propodeum as in A. radiatus but with sculpturing slightly 
coarser; dorsal area of propodeum with carinae more commonly anastomos- 
ing or rugose. (16) Wing as in A. radiatus. (17) Tegula as in A. radiatus 
6 but much paler. (18-20) Fore, middle and hind legs as in A. radiatus. 
Metasoma: (21) Terga as in A. radiatus. (22) Sterna as in A. radiatus 
but darker and never with metallic tints on sternum 4. 

Agapostemon heterurus Cockerell 

Agapostemon heterurus Cockerell 1917 (a). Type 9, U.S. National Museum. 

In addition to the type and paratype, I have seen a third specimen of A. 
heterurus. All three of these specimens are females; all are in the U.S. 
National Museum; all are labeled, "Palcazul Peru//Collection/ Rosenberg." 

Diagnosis. The female may be distinguished from most other species by 
the acute lateral angle and posterior lobe of its pronotum; from A. mourei 
and A. boliviensis by the lack of yellow on its clypeus and by its less finely 



484 The University Science Bulletin 

branched mesoscutal pubescence; from A. lanosus by its finer genal striae 
and its dark metallic blue-green head and mesosoma. 

Description. Female as in A. mourei but head and mesosoma dark 
metallic blue to blue-green; clypeus lacking yellow; pubescence on meso- 
and metanotum less dense and woolly; tegulae dark brown; metasomal terga 
often with dark brown bands centrally. 

Agapostemon hispaniolicus n. sp. 

This species is named for the island of Hispaniola where it occurs. 

The male holotype is labeled, "Port-au-Prince/ & vie, Haiti/3-X 1934/ 
Darlington." The allotype is labeled, "Manneville/Haiti/Nov 16-17, '34/ 
Darlington." The types in the Museum of Comparative Zoology, Harvard 
University, Cambridge, Massachusetts. 

Distribution. In addition to the holotype and allotype, I have seen 
specimens from Santo Domingo (2 9) and Samana (1 8 , June) in the 
Dominican Republic; and from La Moriniere (1 2 , March, 125 ft.) and 
Port au Prince (12, April, 300 ft.; 1 $ ) in Haiti. 

Diagnosis. The male may be distinguished from other West Indian spe- 
cies, except A. viriduhts and A. obscuratus, by its black metasomal terga and 
toothed hind femora; from A. obscuratus by its bright metallic blue-green 
head and mesosoma; and from A. viridulus by the contiguous punctures on 
its mesoscutum. The female may be distinguished from other West Indian 
species, except A. viridulus and A. swain sonae, by its bright green head and 
mesosoma and black metasomal terga; from A. viridulus and A. swainsonae 
by the finer sculpturing of its interocular area. 

Description 
male (Figs. 115-116, 148) 
As in A. viridulus but with slightly finer and denser sculpturing; with 
slightly narrower apical groove on basal ridge of basitarsus; and with shal- 
lower and more numerous longitudinal grooves laterally on gonocoxa. 

female (Figs. 37-38) 

As in A. viridulus but bluer, smaller, and more finely sculptured, clypeal 
area not as elongate, and with narrower distal keel on labrum. 



l 6 



Agapostemon inca n. sp. 

This species is named for the Indians of the Quechuan tribe inhabiting 
Peru. 

The male holotype and one male paratype are labeled, "Hda San Juan/ 
Col Perene Peru/ 16 June 1920/ Cornell Univ. Exp.// Cornell Univ. Ex-/ 
pedition Lot 607/ Sub 95." The above two specimens are in the collection 
of Cornell Univ., Ithaca, N.Y. The only other specimen I have seen is a 



Revision of the Bee Genus Agapostemon 485 

male (paratype) in the American Museum of Natural History labeled, 
"PERU: Montenegro/ Bagua, Amazonas/ Sept. 29-Oct. 2, 1963/ 350m. 
Wygodzinsky." 

Diagnosis. The male may be distinguished from those of most other 
species by the presence of a medial "button" at the base of its last visible 
(6th) sternum. This sternal "button" is much longer and broader than that 
of the closely related A. mourei (Figs. 139-140), and the hind femur, tibia 
and basitarsus are more slender than those of A. mourei (Figs. 167-168). 
Agapostemon inca may be distinguished from the related A. intermedins, 
A. boliviensis and A. lanosus by the larger and differently shaped medial 
plate of its gonostylus (Figs. 175, 177 A-B, 192). 

Variation. The general appearance of the males differs strikingly be- 
cause the black markings of the appendages and metasoma may be replaced 
by pale amber markings. 

Description 

male (Figs. 140, 168, 175, 218) 

As in A. mourei but with head and mesosorna dark coppery green; an- 
tenna dark brown to black with yellow spot at base of lower side of scape; 
mesosorna with slightly finer punctures and less coarse rugae; tegula brown 
to brown-black, nearly obscuring anterior, submarginal, yellow crescent; 
fore leg yellow with coxa brown, trochanter brown but for anterior yellow 
spot, and brown on posterior of femur and tibia; middle leg brown with 
yellow spot on anterior of trochanter, yellow streak on anterior of femur, 
and small yellow spot anteriorly at either end of tibia; hind leg (Fig. 168) 
more slender, and dark brown with antero-dorsal yellow patch on femur; 
metasomal sterna brown with yellow on sterna 2-3; proximo-medial projec- 
tion, or "button," on sternum 6 (Fig. 140) very broad, base with proximal 
portion wider than distal portion; genitalia (Figs. 175, 218). 

Agapostemon insularis n. sp. 

The name of this species refers to its island distribution. 

I have seen four specimens (3 $ , 1 9 ) from the island of Hispaniola. 
The female holotype and the allotype are labeled "Aux Cayes, Haiti/Mch. 
15-20, '22/F. 4644." The two male paratypes are from Carrefour, Haiti 
(April) and the San Francisco Mts., Dominican Republic (Sept.). The 
types are in the American Museum of Natural History, New York. 

Diagnosis. The male may be distinguished from other West Indian 
species, except A. jamaicensis and A. poeyi, by its medial metallic tints on 
metasomal tergum 3; from A. jamaicensis and A. poeyi by its lack of postero- 
lateral metallic tints on metasomal tergum 1. The female may be distin- 
guished from other West Indian species, except A. poeyi, by its conspicuous 



486 The University Science Bulletin 

medial metallic green tints on metasomal sterna 3-4; and from A. poeyi 
by its finely punctate mesoscutellum. 

Description 

MALE 

General coloration of head and mesosoma bright metallic green to blue- 
green, metasoma brown with yellow bands and tinted with green posteriorly. 

Head: as in A. vieqitesensis. Mesosoma: as in A. viequesensis but with 
sculpturing slightly finer, pronotum with lateral angle and posterior 
lobe both slightly more angular and with fewer postero-lateral carinulae, 
and tegula with metallic green tints slightly more extensive. Metasoma: 
(21) Terga as in A. columbi but with conspicuous metallic tints on tergum 
3; pygidium rounded apically as in A. viequesensis. (22) Sterna as in A. 
viequesensis. (23) Genitalia as in A. poeyi but with apex of apical stylus 
on gonostylus slightly smaller (much closer to that of A. poeyi than that 
of A. viequesensis). 

FEMALE 

General coloration of head, mesosoma and metasoma bright metallic 
greenish blue. Head: pubescence as in A. viequesensis. (1-4) Labrum, 
clypeus, interocular area and vertex as in A. viequesensis. (5) Gena much 
more coarsely carinate than that of A. viequesensis but not quite as coarse 
as in A. poeyi. (6) Malar area absent. (7-8) Mandible and antenna as in 
A. poeyi. Mesosoma: pubescence as in A. viequesensis but slightly more 
fuliginous on mesonotum and metanotum. (9) Pronotum as in A. vieque- 
sensis but with lateral angle and posterior lobe both more angular and with 
fewer and coarser (not as coarse as in A. poeyi) horizontal carinae postero- 
laterally. (10-12) Mesoscutum, mesoscutellum and metanotum as in A. 
viequesensis. (13-14) Mesepisternum and metepi sternum as in A. vieque- 
sensis but with sculpturing very slightly coarser. (15) Propodeum as in 
A. poeyi but with finer sculpturing (not quite as fine as in A. viequesensis). 
(16-17) Wing and tegula as in A. viequesensis but tegula darker with metal- 
lic tints more extensive. (18-20) Fore, middle and hind legs dark brown, 
coxae with strong metallic tints. Pubescence dark amber to brown, scopal 
hairs white. Metasoma: as in A. viequesensis but with terga wholly 
metallic, and sterna 3-4 with strong metallic tints. 

Agapostemon intermedins n. sp. 

The name refers to the phenotype of this species which is intermediate 
between that of A. semimelleus and that of A. heterurus. 

The male holotype, the allotype, and the 28 paratypes (4 $ , 24 $ ) are 
all from Costa Rica: Gromaco, 34 km southeast of Potrero Grande, on the 



Revision of the Bee Genus Agapostemon 487 

Rio Coto Brus, southern Puntarenas Province, 21 July 1963, 1,000 ft. (C. D. 
Michener and W. Kerfoot) and are in the Snow Entomological Museum 
at the University of Kansas. 

Distribution. In addition to the types I have seen specimens from the 
following localities: Turrialba (2 <$ , 8 July; 1 $ , 31 July; 15,1 Aug.; 
15,3 Aug.; 1 S , 10 Aug.; 1 S , 16 Aug.; 2 $ , 28 Aug.; 1 9 , 12 Sept.), 
4.5 mi. SW Cartago (19,3 July) and Pozo Azul at the confluence of Rios 
Parrita and Candelaria (19,9 Dec), Costa Rica; 10 miles east of Cumana, 
Venezuela (2 9 , 13 March); Macas, Rio Upano (1 $, 25 Jan.) and 
Pichilingue, Los Rios (59,2 Feb.), Ecuador; Monson Valley, Tingo Maria, 
Peru (1 5, 2 Nov.; 2 6, 21 Nov.; 3 S, 29 Nov.). 

Diagnosis. The male is easily distinguished from all other North and 
Central American Agapostemon by its medial "button" (flattened process 
resembling human tongue and attached to sternum in much the same way as 
human tongue is attached to floor of mouth) at the base of the last visible 
(6th) sternum (Fig. 136), and from the South American Agapostemon with 
similar "buttons" on sternum 6 by the low basal ridge and inconspicuous 
apical groove ventrally on its slender basitarsus (Fig. 151). The female may 
be distinguished from the other North and Central American species of 
Agapostemon by the basal yellowish bands on its metasomal terga 2-5 (may 
be obscured by overlapping terga posteriorly), and from South American 
species with yellow metasomal bands by its bright metallic green or coppery 
head and mesosoma and the unbroken yellow band on its clypeus (the 
yellow-banded South American females have only faint metallic tints and 
nearly always have the yellow clypeal band broken medially). 

Variation. The coloration of females of this species varies considerably. 
While the female head and mesosoma are usually bright metallic green with 
amber and fuliginous pubescence, I have collected a female from Turrialba, 
Costa Rica, with bright coppery head and mesosoma and with fulvus pubes- 
cence. At first I was inclined to regard it as a separate species but the lack 
of morphological differences, and the slightly fulvus pubescence of the other- 
wise typical males from Turrialba, together with the similar pattern of varia- 
tion in the sympatric A. leunculus, lead me to regard these differences as 
geographic variation. 

The color of the yellow-banded metasomal terga of females varies from 
pale amber (Gromaco and Pozo Azul, Costa Rica; Cumana, Venezuela; and 
Pichilingue, Ecuador) to brown-black (Turrialba, Costa Rica; Cumana, 
Venezuela; and Pichilingue, Ecuador). Although it is difficult to assess 
such variation on the basis of so few specimens it appears that the color of 
the yellow-banded metasomal terga varies both within and among popula- 
tions as does that of the North American A. melliventris. 



488 The University Science Bulletin 

Description 
male (Figs. 81-82, 136, 151, 192, 214) 

General coloration of head and mesosoma bright metallic green, meta- 
soma banded with black and yellow. Head (Figs. 81-82) : with pubescence 
golden to fulvus on interocular area and vertex, becoming creamy on gena. 
(1) Labrum as in A. texanus but with apical portion much shorter and more 
acute at apex. (2) Clypens with large, scattered, shallow punctures on 
lower (yellow) Y- 1 /^ becoming contiguous on upper (green) Y}- l /i. (3) 
Interocular area with small to moderately small, deep, contiguous punctures 
below ocelli, becoming shallower and medium-sized at level of antennal 
sockets and still shallower below antennae; supraclypeal protuberance with 
large contiguous moderately deep punctures. Pubescence between antennal 
sockets Yi- 7 /} as long as scape, becoming shorter laterally, dorsally and ven- 
trally and subtomentose ventro-laterally. (4) Vertex with small, deep, con- 
tiguous punctures anteriorly and laterally, becoming shallower and slightly 
larger between ocelli and transversely carinate posteriorly. (5) Gena with 
parallel, subcontiguous, short (often very weak) carinae extending dorso- 
lateral^ from antero-ventral margin. Pubescence about as long as scape 
postero-medially, becoming shorter anteriorly, laterally and dorsally. (6) 
Malar area yellow, very short, and with minute, appressed hairs. (7) 
Mandible yellow with apical ! / 3 ferruginous to amber. (8) Antenna brown 
to brown-black above, scape yellow below, pedicel and flagellum yellowish 
to amber below. Mesosoma: pubescence creamy or yellowish becoming 
golden to fulvus on mesonotum and metanotum. (9) Pronotum with lateral 
angle and posterior lobe acutely pointed, lateral angle projecting antero- 
lateral^ ; with conspicuous sharp carina extending postero-ventrally from 
apex of lateral angle; with weak irregular rugae postero-laterally. Pubescence 
above lateral angle and posterior lobe golden to fulvus, yellowish to creamy 
below. (10) Mesoscutum with large conspicuous flange laterally on anterior 
margin extending antero-dorsally; punctures small, very deep and contiguous, 
becoming rugose antero-laterally. (11) Mesoscutellum with medial punc- 
tures like those of mesoscutum, becoming slightly larger and less dense 
laterally; relatively shiny area with widely separated punctures on each side 
of midline. (12) Metanotum coarsely punctate to rugose medially, becoming 
moderately coarsely rugose laterally. (13) Mesepistemum moderately 
coarsely rugose anteriorly, becoming finely rugose posteriorly. (14) Metepi- 
sternum moderately finely rugose with horizontal rugae longer and slightly 
more prominent than other rugae, or with irregular horizontal carinae. (15) 
Propodeum with propodeal carina usually strong and conspicuous; propodeal 
shield very coarsely and deeply punctate to very coarsely rugose; dorsal area 
coarsely and very deeply rugose anteriorly and medially, becoming deeply 
and moderately rugose to coarsely punctate laterally; lateral area with small 



Revision of the Bee Genus Agapostemon 489 

and very deep subcontiguous punctures anteriorly and centrally, becoming 
coarsely rugose adjacent to propodeal carina. (16) Wing hyaline or nearly 
hyaline, veins dark brown. (17) Tegula pale to dark transparent amber with 
broad submarginal yellow band anteriorly (partially obscured by pubes- 
cence) and amber to dark brown on proximal margin. (18) Fore leg yellow 
with brown streaks posteriorly on femur and tibia. (19) Middle leg yellow 
with coxa brown and with brown postero-dorsally on tibia and postero- 
dorsally at apex of femur (also present posteriorly on trochanter of some 
specimens). (20) Hind leg (Fig. 151) yellow with upper x / 2 of coxa brown 
and tinted metallic green; with trochanter brown basally, becoming amber 
apically; with brown apically on femur and antero-dorsally on tibia. 

Metasoma: (21) Terga 2-6 dark brown to brown-black on posterior l / 2 
and yellow on anterior l / 2 (often partially concealed by overlap), tergum 
1 dark brown to brown-black with broad yellow band antero-dorsally (cen- 
trally) ; very conspicuously punctate with subcontiguous punctures largest 
on tergum 1, becoming progressively smaller on posterior terga. Pubescence 
golden to fulvus anteriorly on tergum 1, laterally on terga 1-7 and dorsally 
on terga 5-7; very short dorsally on terga 1-4, golden to fulvus on yellow 
bands and dark brown to brown-black on dark bands. (22) Sterna yellow 
with brown on sternum 1 and with transverse bands on posterior margins 
of sterna 2-5 narrowest on sternum 2, becoming progressively wider (espe- 
cially medially) posteriorly; sternum 6 with proximo-medial projection, or 
"button," in form of human tongue (Fig. 136) ; pubescence limited to scat- 
tered, moderately long, fulvus to golden hairs on exposed areas. (23) 
Genitalia (Figs. 192, 214) with ventral lobes very short and with a brush 
of large stiff hairs. 

female (Figs. 47-48) 

General coloration of head and mesosoma bright metallic green or cop- 
pery, metasoma pale amber to black and with yellow bands. Head (Figs. 
47-48) : pubescence golden to fulvus, becoming creamy to golden on gena. 
(1) Labrum as in A. texanus. (2) Clypeus with large scattered punctures 
on lower l / 2 , becoming subcontiguous on upper l / 2 . (3) Interocular area 
with small, moderately deep, contiguous punctures below ocelli, becoming 
rugose from just above antennal sockets to lower margin; supraclypeal pro- 
tuberance with medium sized punctures separated by nearly twice their 
diameters and with short, inconspicuous horizontal rugae. (4) Vertex with 
small, moderately deep, contiguous punctures anteriorly and laterally, be- 
coming subcontiguous between ocelli and transversely carinate posteriorly. 
(5) Gena with fine, contiguous, parallel carinae extending postero-dorsally 
from antero-ventral margin. (6) Malar area amber, very short to absent. 
(7) Mandible yellow, becoming dark amber or ferruginous on apical ! /3- 



490 The University Science Bulletin 

(8) Antenna dark brown to brown-black with underside of flagellum 
amber. Mesosoma: pubescence creamy to golden, becoming golden (with 
large scattered hairs dark) to fulvus (with large scattered hairs pale) 
on mesonotum and metanotum. (9) Pronotum as in $ but with weak 
carina extending from lateral angle to posterior lobe. (10) Mesoscutum as 
in $ but with most punctures slightly smaller and shallower (large scattered 
hairs in large punctures). (11) Mesoscittellum with sculpturing like that of 
mesoscutum but with small punctures slightly shallower. (12) Metanotum 
moderately finely rugose. (13) Mesepisternum as in $ but with sculpturing 
slightly more coarse. (14) Mete pi stern um horizontally carinulate with inter- 
connecting rugae. (15) Propodettm with strong propodeal carina; propodeal 
shield with conspicuous carinae extending dorso-laterally from medial groove 
and with fine rugae between conspicuous carinae; dorsal area coarsely and 
extremely deeply rugose anteriorly and medially, becoming moderately 
deeply and less coarsely rugose laterally; lateral area horizontally carinulate 
with small scattered punctures anteriorly becoming very coarsely and 
horizontally carinate posteriorly. (16) Wing as in S . (17) Tegula as in 
S but slightly darker. (18) Fore leg with dark brown on coxa and tro- 
chanter; femur dark brown, becoming yellow anteriorly and dorsally on 
apical l / 3 ; tibia and tarsus yellowish to pale amber anteriorly, becoming 
brown posteriorly. Pubescence creamy on coxa, trochanter and femur; 
golden on tibia and tarsus. (19) Middle leg dark brown, becoming pale 
amber to yellowish anteriorly and dorsally on apical % of femur and at base 
of tibia; pubescence creamy on coxa and trochanter, becoming golden on 
femur, tibia and tarsus, and dark brown dorsally on tibia and tarsus. (20) 
Hind leg dark brown to pale brown with metallic green dorsally on coxa. 
Pubescence creamy on coxa, trochanter and femur; golden ventrally and 
dark brown to brown-black dorsally on tibia and tarsus. Metasoma: 
(21) Terga 2-5 pale amber to black, with yellow band on basal l / 3 of each 
(may be partially obscured by overlap) ; tergum 1 pale amber to black with 
inconspicuous pale spot dorso-laterally. Punctures small, separated by about 
Y 2 their diameters. Pubescence moderately short, amber to brown dorsally 
on terga 1-4 but with sparsely scattered longer and darker hairs increasing 
in size and abundance posteriorly from tergum 1; medium length, creamy 
to amber anteriorly on tergum 1 and laterally on terga 1-4; moderately long, 
dense and dark brown to black on terga 5-6. (22) Sterna dark brown on 
specimens with brown to brown-black on tergum 1, but sterna 1-3 and 
base of sternum 4 amber on specimens with tergum 1 amber; metallic tints 
basally on sternum 1. Long amber pubescence scattered on exposed areas. 

Agapostemon jamaicensis n. sp. 

This species is named for the island where it occurs. 

I have seen a single male specimen labeled, "Moneage, Jamaica./Sept. 14, 



Revision of the Bee Genus Agapostemon 491 

1917 (497)/Harold Morrison." The holotype is in the United States Na- 
tional Museum, Washington, D.C. 

Diagnosis. The male may be distinguished from other West Indian 
species, except A. poeyi and A. insularis, by its posterior metasomal terga 
strongly tinted with metallic green medially and by its brown eyes and 
translucent amber to brown pterostigma; from A. insularis by its metallic 
green tints postero-laterally on metasomal tergum 1; and from A. poeyi 
by its more inflated hind femur and tibia. 

Description 

male (Fig. 146) 

General coloration of head and mesosoma bright metallic blue-green, 
metasoma dark brown banded with yellow anteriorly, becoming bright 
metallic green posteriorly. Head: as in A. viequesensis but with sculp- 
turing very slightly coarser and with pubescence pale amber. Mesosoma: 
pubescence as in A. viequesensis. (9) Pronotum as in A. viequesensis but 
with lateral angle and posterior lobe slightly more angular. (10-12) Meso- 
scutum, mesoscutellum and metanotum as in A. viequesensis. (13-14) 
Mesepisternum and metepistemum with sculpturing coarser than in A. 
viequesensis but not as coarse as in A. poeyi. (15) Propodeum as in A. poeyi 
females but with sculpturing slightly finer and antero-lateral area with 
large punctures replacing carinae. (16-19) Wing, tegula, fore and middle 
legs as in A. viequesensis. (20) Hind leg as in A. viequesensis but with 
femur and tibia more inflated, apical groove on basitarsus broader and 
convex centrally, and basal ridge more prominent with broad flat area on 
apical portion (Fig. 146). Metasoma: (21) Terga as in A. viequesensis 
but slightly darker and with bright metallic green replacing brown band 
completely on tergum 4 and partially on terga 1-3 and 5-6. (22) Sterna as in 
A. viequesensis but dark amber. (23) Genitalia as in A. poeyi but with 
apex of apical stylus of gonostylus more slender than that of A. poeyi (not as 
slender as A. viequesensis). 

Agapostemon kohliellus (Vachal) 

Halictus (Agapostemon) kohliellus Vachal 1903. Type 9, Naturhistorisches Museum, Vienna. 

I have examined the type of H. kohliellus and the locality label is illegible. 
This species is atypical of the genus inasmuch as the labrum of the female 
is trilobed (Fig. 31). For the present study it is included in Agapostemon 
because it is phenetically closer to this genus than to any other. Ultimately 
it may be desirable to erect a separate genus for this species. 

Distribution. Other than the type, I have seen specimens from Port au 
Prince (1 9, March) and Cul-de-Sac Plain (3 S, Dec.) Haiti and from 
Rio Toa Baracoa, Cuba (1 S , April). 



492 The University Science Bulletin 

Diagnosis. The male may be distinguished from all other West Indian 
species by its truncate clypeus, hind femur lacking tooth, and unique hind 
basitarsus. The females may be distinguished from all other West Indian 
species by the yellow band on the clypeus and by the yellow bands on the 
honey-colored metasoma. 

Description 

male (Figs. 109-110, 141, 191, 212) 

General coloration of head and mesosoma metallic green, metasoma 
yellow with brown bands. Head (Figs. 109-110): pubescence white to 
yellowish on clypeus, interocular area and vertex; white on gena. (1) 
Labrum yellow, with apical margin transparent; apex truncate to very 
slightly bilobed, basal x /i without medial depression on transverse basal ridge; 
labral surface shiny and impunctate with row of small, simple submarginal 
setae. (2) Clypeus yellow with metallic green on upper margin, truncate, 
but not nearly as much as in A. nasutus (Figs. 85-87); flattened lower por- 
tion with several small shallow punctures and nearly glabrous, narrow upper 
portion with shallow scattered punctures and with short sparse pubescence. 
(3) Interocular area with deep, coarse, contiguous punctures above; deeply 
rugulose around antennal sockets, becoming less deeply rugulose below. 
Supraclypeal protuberance small but relatively acute, shallowly and coarsely 
rugose below, weakly rugulose above. Pubescence between antennal sockets 
about % as long as scape, becoming slightly shorter dorsally, laterally and 
ventrally; becoming short appressed tomentum ventro-laterally. (4) Vertex 
with deep, contiguous, medium-sized punctures anteriorly, laterally and 
between ocelli; becoming rugose posteriorly. Pubescence like that of dorsal 
portion of interocular area. (5) Gena with weak rugae extending postero- 
dorsally from antero-ventral margin; postero-medial pubescence slightly 
longer than that between antennal sockets, becoming shorter anteriorly, 
laterally and dorsally. (6) Malar area absent. (7) Mandible yellow, ferru- 
ginous on apical l /?,. Row of hairs on postero-ventral margin only l /^ as long 
distally as proximally. (8) Antenna brown above, but scape with yellow on 
basal x /i of upper surface; scape and pedicel yellow below, flagellum amber 
below. Mesosoma: pubescence white, yellowish on mesonotum and 
metanotum. (9) Propodeum with lateral angle and posterior lobe rounded; 
sculpturing weak and inconspicuous. (10) Mesoscutum with deep, small, 
contiguous punctures becoming rugose antero-laterally. (11) Mesoscutellum 
with small, deep punctures separated by at least their diameters, becoming 
rugose laterally. (12) Metanotum shallowly rugose with small, deep punc- 
tures more abundant than those of mesoscutellum. (13) Mesepisternum 
finely rugulose anteriorly, becoming very finely rugulose posteriorly. (14) 
Metepisternum horizontally carinulate. (15) Propodeum rounded posteriorly 



Revision of the Bee Genus Agapostemon 493 

with propodeal carina absent; small, deep punctures laterally separated by 
at least their own diameters, becoming rugulose or contiguously punctate 
postero-dorsally; propodeal shield shallowly rugulose. (16) Wing trans- 
parent; veins and pterostigma dark brown, radius almost black. (17) Tegitla 
pale transparent honey-colored with yellow on basal margin and with yellow 
band almost reaching distal margin. (18-19) Fore and middle legs yellow 
with coxae metallic green. (20) Hind leg (Fig. 141) yellow with trochanter 
brownish, brown apically on dorsal surface of femur and amber on all but 
ventral surface of tibia, and with coxa metallic green; femur without tooth; 
femur and tibia moderately inflated; basitarsus broad but flat with basal 
ridge narrow and gently curved apically; basitarsus lacking apical groove. 
Metasoma: (21) Terga. Tergum 1 yellow, amber anteriorly and brown 
apically; terga 2-6 with basal yellow band and apical brown band of equal 
width; tergum 7 amber. Pubescence minute and inconspicuous dorsally, 
amber on yellow bands, and black on brown bands; long brown hairs 
dorsally on terga 5-7; long pale hairs ventro-laterally on terga 4-6 and 
anteriorly on tergum 1. (22) Sterna amber with white hairs on exposed 
areas. (23) Genitalia (Figs. 191, 212), ventral lobe short with long brush 
of hairs on short inner portion. 

female (Figs. 31-32) 

General coloration of head and mesosoma bright metallic green, meta- 
soma honey-colored with yellow bands. Head (Figs. 31-32): with short 
white pubescence. (1) Labrum (Figs. 31A-B) with acute lateral lobes on 
apical l / 2 . (2) Clypeus yellow below, green above; yellow portion shiny, 
sparsely punctate, green portion with medium-sized deep, contiguous punc- 
tures. (3) Interocular area rugulose, becoming finely and deeply punctate 
above; supraclypeal protuberance low and gently rounded with medium- 
sized punctures separated by their own diameters and with numerous 
horizontal rugae; pubescence between antennal sockets only about V3 as 
long as scape. (4) Vertex with fine, deep, contiguous punctures anteriorly, 
becoming smaller, shallower and subcontiguous laterally and faintly rugulose 
posteriorly; area between ocelli shiny with small scattered punctures. (5) 
Gena with fine, contiguous rugulae extending postero-dorsally from antero- 
ventral margin; pubescence postero-medially about 1.5 X as long as pubes- 
cence between antennal sockets and becoming shorter anteriorly, laterally 
and dorsally. (6) Malar area absent. (7) Mandible yellow with apical J/3 
ferruginous; single row of hairs on postero-ventral margin long basally, 
becoming Y 3 as long distally. (<S) Antenna brown, flagellum yellowish to 
amber below. Mesosoma: short white to yellowish pubescence. (9) 
Pronotum with lateral angle and posterior lobe rounded; sculpturing very 
fine, weak and irregularly rugulose ventro-laterally, becoming smooth dorso- 



494 The University Science Bulletin 

medially. (10) Mesoscutum with small contiguous punctures, becoming 
faintly rugulose antero-laterally. (11) Mesoscutellum shiny, with scattered 
large punctures and more numerous very small punctures. (12) Metanotum 
sparsely punctate, irregularly and transversely rugulose. (13) Mesepistemum 
with shallow, fine rugae anteriorly, becoming much finer and transversely 
rugulose to carinulate posteriorly. (14) Metepi sternum with shallow, regular 
transverse carinulae. (15) Propodeum with distinct but weak propodeal 
carina; very fine, regular, horizontal carinulae laterally and on propodeal 
shield; dorsal area with very fine carinulae extending postero-laterally from 
anterior margin. (16-17) Wing and tegula as in $ . (18-20) Fore, middle 
and hind legs yellowish to amber with white to yellowish pubescence be- 
coming fuliginous dorsally on middle and hind tibiae. Metasoma: (21) 
Terga amber to pale brown with yellow bands basally on terga 2-5; tergum 
1 may have pale yellow band anteriorly on dorsal portion interrupted 
medially; pubescence as in 6 . (22) Sterna amber to brown with long white 
hairs scattered on exposed areas. 

Agapostemon lanosus n. sp. 

The name of the species refers to the woolly appearance of the mesoscutal 
pubescence. 

The male holotype and the allotype are labeled, "PERU:/ Monson 
Valley/ Tingo Maria/ XI-21-1954// E. I. Schlinger/ &E. S. Ross collectors." 
The two male paratypes were collected by Ross and Schlinger from the 
same locality on "29-XI-1954" and "XII-23-1954." At present, these types 
are the only known specimens and they are in the California Academy of 
Sciences, San Francisco. 

Diagnosis. The male may be distinguished from most other species by 
the presence of a medial "button" at the base of its last visible (6th) sternum 
(Fig. 138). It may be distinguished from other species with sternal "buttons" 
by the presence of 2-3 long (about twice the length of adjacent setae), thick- 
ened, apically hooked setae laterally on the posterior margin of the antepenul- 
timate visible (4th) sternum. 

The female may be distinguished from most other species by the acute 
lateral angle and posterior lobe of its pronotum and by the dense woolly pu- 
bescence of its mesoscutum; from A. moitrei and A. boliviensis by the lack of 
yellow on its clypeus, by its less finely branched mesoscutal pubescence and 
by its coarser genal sculpturing; from A. heterurus by the parallel striae 
dorsally on its pronotum and by the coppery luster of its head and mesosoma. 

Variation. The general appearance of the males varies strikingly be- 
cause the black markings of the appendages and metasoma may be replaced 
by pale amber markings. 



Revision of the Bee Genus Agapostemon 495 

Description 
male (Figs. 138, 169, 177B, 216) 

As in A. mourei but with head and mesosoma bright coppery green; 
antenna dark brown to black with scape yellow below; mesoscutellum with 
punctures widely scattered in central region; tegula darker amber with an- 
terior, submarginal, yellow crescent partially obscured proximally; fore leg 
yellow with brown on posterior of coxa, trochanter, femur and tibia and 
anteriorly at base of coxa; hind basitarsus (Fig. 169) yellow with dark 
brown to black on ventral carina; metasomal sterna dark brown with yellow 
on sterna 2-3; sternum 6 (Fig. 138) with proximo-medial projection, or 
"button," with each lateral flange slightly less than l / 2 width of base (as 
measured on transverse section through widest part of "button") ; genitalia 
(Figs. 177B, 216). 

FEMALE 

As in A. mourei but clypeus without yellow band; gena with deeper 
striae; meso- and metanotum with pubescence darker and slightly less 
woolly; propodeum dorsally with coarse, deep striae extending postero- 
laterally from anterior margin; metasoma dark brown with yellowish amber 
on tergum 1 and proximo-laterally on terga 2-5. 

Agapostemon leunculus Vachal 

Agapostemon leunculus Vachal 1903. Lectotype 2, Museum National D'Histoire Naturelle, 

Paris. 
Agapostemon vulpicolor Crawford 1906 (new synonymy). Type 2, U.S. National Museum. 
Agapostemon nasutus var. ater Friese 1916 (new synonymy). Type 5, whereabouts unknown. 

From the 30 syntypes of Agapostemon leunculus in the Paris Museum, 
I have designated a female as lectoholotype. I have seen the type of A. 
vulpicolor but not that of A. nasutus ater. As noted by Michener (1954), 
A. nasutus ater is not conspecific with A. nasutus. I have seen a female 
"Typus" (not a type, but identified by Friese) of A. nasutus ater from the 
U.S. National Museum labeled, "Costa Rica,/San Jose/1903." Friese did 
not describe the male, probably because he was expecting to find something 
with the "hog-nosed" aspect of A. nasutus. The only difference between 
A. leunculus and A. nasutus ater is the coloration of the pubescence and 
integument of the head and the dorsal surface of the thorax. I have not 
found any differences among the males that can be correlated with the varia- 
tion among females. Because I have seen females intermediate in coloration 
(see discussion of variation) I am considering the difference between A. 
leunculus and A. nasutus ater as local variation. 

Distribution. Agapostemon leunculus is found in Mexico exclusive of 
the central plateau, as far north along the eastern coast as ten miles south- 
west of Pharr, Texas (1 2 collected by C. D. Michener) and as far north 



496 The University Science Bulletin 

along the western coast as Rosario, Sinaloa. It has been collected in the 
mountainous regions of southern Mexico but has not been taken on the 
Yucatan Peninsula. It has also been collected in the mountains of El 
Salvador, Honduras, Costa Rica, Panama, Colombia and Ecuador (Fig. 5). 
Although information on altitudinal distribution is scant, this species has 
not been found at low altitudes (less than 1,000 m) in the hot tropics. It has 
been collected from 32 m (105 ft.) at Rosario, Sinaloa, Mexico, up to 7,100 
ft. (2,164 m) near Nochixtlan, Oaxaca, Mexico; from 3,300 ft. (1,006 m) at 
Santiago de Puriscal, Costa Rica, up to 9,000 ft. (2,743 m) on Volcan Irazu, 
Costa Rica; 7,800 ft. (2,377 m) at Popayan, Colombia, and 2,800 m (9,184 
ft.) at Calacali, Ecuador. A. leunculus has been collected in February and 
October in Ecuador; February in Colombia; every month but January, 
March, June and November in Central America; every month but January 
and November in Mexico; and on March 31 near Pharr, Texas. Undoubtedly 
this bee is more widely distributed altitudinally and seasonally than is in- 
dicated by these records. I have seen fewer than 300 specimens, and this 
sample is probably biased because bee collectors seldom work as high as 
3,000 m and because most of the specimens were collected by North Ameri- 
cans, who seldom have an opportunity to collect from October through May. 

Diagnosis. The male may be differentiated from males of many other 
species by its slender, untoothed hind femora and rounded lateral pronotal 
angle; from A. nasutus by its normal clypeal region; from A. rhopalocera 
by its normal antennae; from A. peninsidaris and A. mexicanns by its lack 
of enlarged subapical hairs on metasomal sternum 4; and from A. melli- 
ventris by the black on the basal portion of its metasomal tergum 1. The 
female may be distinguished from other species with black metasoma and 
green or coppery head and mesosoma by the prominent carina extending 
postero-ventrally from the rounded lateral angle of its pronotum. 

Variation. The coloration of the head and dorsal mesosomal region, as 
well as the pubescence thereon, is quite variable in females of this species. 
Most of the females are metallic green on the head and thorax. Some of the 
females from Costa Rica (Volcan Irazu, San Jose, Cartago and San Mateo) 
have reddish or coppery integument on the head and the dorsal mesosomal 
area. The pubescence on these areas is orange. Although nowhere else so 
extreme, I have seen females from Mexico (e.g., Orizaba, Vera Cruz; 
Ajijic, Jalisco; Tehuacan, Puebla; and near Oaxaca, Oaxaca) with orange 
pubescence and with coppery tints on the head and the dorsal mesosomal 
area. In still other areas (e.g., near Ciudad del Maiz and San Luis Potosi) 
the integument is green but the pubescence orange. In those regions where 
the orange of the pubescence and the coppery color is most pronounced, the 
yellow band on the clypeus of females is often reduced or absent. As these 
regions are relatively cool and wet, it is not surprising that the females are 



Revision of the Bee Genus Agapostemon 497 

more darkly pigmented. The pigmentation of the males is correlated with 
that of the females. Those males with enlarged black or brown markings 
are found with the darker females. As in the females, intergrades may be 
found between the lightest and darkest males. It is the presnce of color 
intergrades and the lack of structural differences that lead me to synonymize 
A. leunciilus and A. nasutus ater. Some of the females from Costa Rica, 
Colombia and Ecuador have fuliginous pubescence on the head and dorsal 
mesosomal regions, but sympatric intermediates exist for this condition also. 
The tegulae of those females with orange or fuliginous pubescence are 
often dark brown to black, unlike the more common yellow and trans- 
parent tegulae of females from warmer, drier areas. 

Description 

male (Figs. 83-84, 143, 193, 211) 

General coloration of head and mesosoma metallic green, metasoma with 
black and yellow bands. Head (Figs. 83-84): (1) Labrum as in A. texamis 
but more acute apically. (2) Clypeus with small scattered punctures on 
yellow portion, larger and deeper scattered punctures on green portion; 
scattered yellowish hairs directed ventrally. (3) Interocular area with punc- 
tures above antennal sockets contiguous, deep and nearly rugose; coarsely 
and shallowly punctate at level of antennal sockets becoming coarsely, 
shallowly and irregularly rugose below; supraclypeal protuberance with 
coarse, contiguous punctures; pubescence pale amber. (4) Vertex with 
fine, contiguous punctures becoming coarse posteriorly; pubescence pale 
amber. (5) Gena with sparse, shallow carinae extending postero-dorsally 
from antero-ventral margin; pubescence dense, white, longest posteriorly. 
(6-7) Malar area and mandible as in A. texantts. (8) Antenna dark brown 
to black above except basal y 2 of scape; yellow to amber below, basal ! / 2 of 
scape usually yellow. Mesosoma: pubescence moderately long, evenly 
distributed (except pronotum), and whitish to amber. (9) Pronotutn with 
weak carina running postero-ventrally from lateral angle; condition of lateral 
angle and posterior lobe intermediate between A. nasutus and A. texanus. 
Pubescence above and between lateral angle and posterior lobe similar to 
that of mesoscutum; short, white, appressed, inconspicuous pubescence on 
postero-ventral surface. (10) Mesoscutum with punctation as in A. nasutus 
but slightly finer and with rugosity restricted to antero-lateral portions. 
(11) Mesoscutellum with punctures slightly deeper and larger than those 
of mesoscutum. (12) Metanotum coarsely punctate medially to shallowly 
rugose laterally. (13) Mesepisternum with punctation as in A. nasutus. 
(14) Metepisternum with predominately horizontal, irregularly anas- 
tomosing rugae not as regular as in A. nasutus. (15) Propodeum coarsely 
punctate antero-laterally becoming moderately rugose postero-laterally and 



498 The University Science Bulletin 

very coarsely rugose dorsally; propodeal carina weak, evenly rounded dor- 
sally, sometimes obscured by coarse rugae; propodeal shield very coarsely 
and irregularly rugose. (16) Wing transparent, not darkened apically, radius 
scarcely darker than other veins. (17) Tegula as in A. nasutus but trans- 
parent brown proximally and with less yellow. (18) Fore leg yellow, often 
with postero-dorsal black streak on any or all segments but tarsus; pubes- 
cence as in A. nasutus but yellow to amber on tibia and tarsus. (19) Middle 
leg yellow with coxa brown to black, postero-dorsal brown streak on 
proximal l / 2 of tibia, and commonly with brown streak on part or entire 
length of postero-dorsal surface; pubescence like that of fore leg but shorter 
on coxa, trochanter and femur. (20) Hind leg (Fig. 143) coxa metallic green 
and usually with some yellow on ventral surface; trochanter predominately 
yellow, brownish dorsally to wholly brown or black; femur with brown 
dorsally and posteriorly at apex; proximal l / 2 of tibia with brown streak on 
dorsal x / 2 sometimes extending to apex; tarsus yellow to brown; pubescence 
as in A. nasutus but slightly shorter and darker. Metasoma: (21) Terga 
shiny; 2-6 with yellow on basal l / 2 , apical 2 / 3 black; tergum 1 black with 
yellow band on antero-dorsal surface; tergum 7 yellow basally, pygidial 
region brown to black; pubescence sparse, simple, moderately long, and 
white on vertical portion of tergum 1; terga 2-4 with short simple hairs, 
white on yellow bands and black on dark bands; terga 5-7 with moderately 
long, stiff, simple, black hairs. (22) Sterna 2-5 yellow with brown bands 
subapically; sternum 1 green basally, apical l / 2 yellow with amber medial 
stripe; pubescence white, short and simple. (23) Genitalia (Figs. 193, 211) 
very simple; penis valves broad and without conspicuous prominences; 
gonostylus elongate with short curved apical stylus and long medial plate of 
moderate width; ventral lobes notched as in A. nasutus but smaller and with 
smaller notches. 

female (Figs. 45-46, 135) 

General coloration of head and mesosoma dark metallic green to blue- 
green, metasoma shiny black with narrow, inconspicuous basal bands of 
white tomentum. Head (Figs. 45-46): (1-4) Labrum, clypeus, interocular 
area and vertex as in A. nasutus but with slightly coarser sculpturing and 
with pubescence white to yellow or fuliginous. (5) Gena with deep, coarse, 
parallel carinae running postero-dorsally from antero-ventral margin ir- 
regularly anastomosing on some specimens; pubescence white (never yellow) 
and shorter than in A. nasutus. (6) Malar area ferruginous, very short. 
(7) Mandible as in A. nasutus. (8) Antenna black above, dark brown be- 
low, undersides of first two flagellomeres often lighter than others. Meso- 
soma: metallic green, never with yellow spot on posterior lobe of prono- 
tum, as in A. nasutus. (9) Pronotum (Fig. 135) as in A. nasutus but with 



Revision of the Bee Genus Agapostemon 499 

prominent carina running postero-ventrally from lateral angle. (10-12) 
Me so scut um, mesoscuteUum and metanotum with punctation as in A. 
nasutus but slightly finer and without shiny lateral area on mesoscuteUum; 
pubescence more dense than in A. nasutus and white to orange or fuliginous. 
(13) Mesepisternum very coarsely rugose anteriorly to moderately rugose 
posteriorly; white to yellowish pubescence. (14) Metepisternum with 
moderately coarse horizontal carinae anteriorly, becoming rugose pos- 
teriorly; pubescence creamy. (15) Propodeum coarsely rugose dorsally and 
postero-laterally, becoming finely rugose antero-laterally; propodeal carina 
prominent, strongly curved dorsally and laterally; propodeal shield with 
irregularly anastomosing carinulae extending laterally from mid-line. (16) 
Wing as in B . (17) Tegula as in £ but may also be brown to black. (18) 
Fore leg dark brown, usually with some yellow on antero-apical portion of 
femur and antero-basal portion of tibia; pubescence yellowish brown ven- 
trally, brown to black dorsally. (19-20) Middle and hind legs dark brown, 
with pubescence brown ventrally and black dorsally. Metasoma: (21) 
Terga black, shinier than in A. nasutus. Narrow basal bands of white 
tomentum on terga 2-5; long, pale hairs scattered on vertical portion of 
tergum 1 and lateral margins of terga 2-5; long black hairs on apical half 
of terga 4-5. (22) Sterna black with long pale hairs on exposed areas. 

Agapostemon melliventris Cresson 

Agapostemon melliventris Cresson 1874 (not 1875). Type $, Academy of Natural Sciences, 

Philadelphia. 
Agapostemon fasciatus Crawford 1901 (new synonymy). Type $, U.S. National Museum. 
Halictus (Agapostemon) plurifasciatus Vachal 1903 (new synonymy). 
Agapostemon digueti Cockerell 1924. Type $, California Academy of Sciences, San Francisco. 

Cresson described this species as new in two different publications (1874, 
1875). Obviously the earlier publication is valid and the latter superfluous. 

I have seen the types of Agapostemon melliventris, A. fasciatus and A. 
digueti and believe they are conspecific. Agapostemon fasciatus Crawford 
has been considered by most authors to be of subspecific rank because it 
differs from A. melliventris sensu stricto in having females with black meta- 
somal terga. I have accorded it neither subspecific status, because in many 
areas (cf. Arizona, Fig. 6) it is sympatric with A. melliventris s.s., nor 
specific status because intermediate forms are too common (cf. discussion 
of variation). 

When Vachal placed Agapostemon in the genus Halictus, Agapostemon 
fasciatus became a junior secondary homonym of Halictus fasciatus Ny- 
lander. Thus Vachal renamed the former Halictus (Agapostemon) pluri- 
fasciatus. Few workers agreed with Vachal's placement of Agapostemon 
as a subgenus of Halictus. If other workers wish to apply a name to this 
dark color form they should use the epithet proposed by Vachal. Cockerell 



500 The University Science Bulletin 

(1937a) and Michener {In Muesebeck, et al., 1951) believed Crawford's 
epithet should be used, but according to the International Code of Zoological 
Nomenclature (anonymous, 1964) secondary homonyms rejected before 
1961 cannot be revived (cf. Agapostemon coloradinus). 

Diagnosis. The male may be distinguished from other species of 
Agapostemon by its yellow metasomal terga with brown to brown-black 
restricted to narrow subapical bands. The female may be distinguished 
from other North American species (except A. nasutus, A. leuncitlus and 
A. peninsitlaris) by the yellow band on the lower margin of its clypeus; 
from A. nasutus by the lack of a yellow spot on the posterior lobe of its 
pronotum; from A. leuncitlus by its lack of a prominent carina extending 
postero-ventrally from the lateral angle of the pronotum; and from A. 
peninsitlaris by its coarser propodeal sculpturing and less dense pubescence. 
In most females of A. melliventris the metasomal terga are pale amber, 
unlike the black metasomal terga of A. peninsitlaris (and some specimens 
of A. melliventris). 

Variation. The most striking variation is the color of the metasoma 
of females (see map, Fig. 6). In females from the central plains of Okla- 
homa, Kansas, Nebraska and South Dakota, the metasoma is almost always 
black. From the opposite side of the range, in females from Baja California 
and Sonora, it is more commonly black than amber. In the remaining por- 
tion of the range the female metasoma is usually amber, but a few females 
with the metasoma black or dark brown may be found in these regions. 
I have seen specimens intermediate with respect to color of abdomen in all 
areas, but they are rare in the central plains. The amount of yellow on the 
clypeus and scape is positively correlated with the paleness of the metasoma, 
those specimens from the central plains having a narrow band of yellow on 
the clypeus and little or no yellow on the scape. 

There is no variation in the color or color pattern of males corresponding 
to that of females. The size and darkness of the brown regions on the meta- 
soma, legs and antennae varies slightly within, but not among, populations. 

One of the most obvious variables in the phenotype is size. As shown 
in Figures 93-94, the linear dimensions of larger specimens may be nearly 
twice those of smaller specimens collected at the same time and place. This 
size variation is somewhat greater among males than among females and 
does not appear to be influenced by geographic factors. 

Description 

male (Figs. 92-94, 155, 188, 203) 

General coloration of head and mesosoma bright metallic green to blue- 
green, metasoma yellow with narrow brown bands; all pubescence white. 



Revision of the Bee Genus Agapostemon 501 

Head (Figs. 92-94): (1) Labrum as in A. texanus but with transparent 
distal margin slightly wider. (2) Clypeus with medium-sized punctures 
separated by about their diameters below, nearly contiguous above. (3) 
Interocular area with medium-sized, moderately deep, contiguous punctures 
becoming minute, shallow and sub-contiguous ventro-laterally. (4) Vertex 
with medium-sized, moderately deep, contiguous punctures anteriorly, be- 
coming slightly smaller and shallower laterally, and rugulose posteriorly 
and between ocelli. (5) Gena with very weak, short, parallel carinulae ex- 
tending postero-dorsally from antero-ventral margin. (6) Malar area yellow; 
short. (7) Mandible yellow with distal third ferruginous to amber. (8) 
Antenna with scape yellow (small brown spot apically on upper side on 
some specimens); pedicel dark brown above, yellow below; flagellum dark 
brown above (yellow at apex of last flagellomere) and yellow to pale amber 
below. Mesosoma: (9) Pronotum with lateral angle and posterior lobe 
rounded; low rounded vertical ridge below lateral angle and faint, parallel, 
horizontal carinae below posterior lobe. (10) Mesoscutum with moderately 
fine, moderately deep, contiguous punctures becoming rugose antero-laterally. 
(11) Mesoscutellum with punctures like those of mesoscutum but slightly 
less dense and sometimes subcontiguous laterally. (12-13) Metanotum and 
mesepisternum moderately finely rugose. (14) Metepistemum with sculp- 
turing as in mesepisternum but slightly coarser and horizontal. (15) Propo- 
deum with propodeal carina moderately weak; propodeal shield coarsely 
but shallowly rugose; dorsal area coarsely and moderately deeply rugose; 
lateral area with moderately coarse rugae postero-dorsally, becoming moder- 
ately finely rugose antero-ventrally. (16) Wing hyaline; radial vein brown, 
other veins and pterostigma pale amber. (17) Tegnla colorless and trans- 
parent with opaque yellow band submarginal anteriorly, marginal basally 
and posteriorly. (18-19) Fore and middle legs yellow with coxae bright 
metallic green. (20) Hind leg (Fig. 155) like fore and middle legs but with 
brown spot distally on dorsal surface of femur and basally on dorsal surface 
of tibia; femur without tooth and basitarsus without apical groove or basal 
ridge. Metasoma: (21) Terga yellow with narrow, subapical, brown 
bands and with apical !4 _1 /3 transparent, pale amber; tergum 1 with anterior 
surface very pale amber to yellow; tergum 7 with pygidium pale amber. 
Short appressed pubescence dorsally becoming moderately long anteriorly, 
posteriorly and laterally. (22) Sterna yellow with faint metallic tints basally 
on sternum 1; narrow brown bands subapically on sterna 2-4; broad brown 
band subapically on sternum 5; and sternum 6 largely brown, with pale 
amber to yellow disto-laterally; moderately short to moderately long scat- 
tered hairs on exposed areas. (23) Genitalia (Figs. 188, 203) with medium- 
sized and slightly elongate ventral lobe on gonocoxite fringed with hairs 
and with short hairs rather dense on posterior surface. 



502 The University Science Bulletin 

female (Figs. 61-62) 

General coloration of head and mesosoma bright metallic green to blue- 
green; metasoma pale amber to black and with broad bands of white 
tomentum. Head (Figs. 61-62): pubescence white, rarely tinged faintly 
with pale yellow on vertex. (1) Labrum as in A. texanus. (2) Clypeus with 
moderately small, scattered punctures below, becoming subcontiguous to 
contiguous above. (3) Interoctilar area rugose below, becoming rugulose to 
coarsely punctate above; supraclypeal area shallowly punctate to weakly 
rugose. (4) Vertex contiguously and coarsely punctate to rugulose anteriorly, 
becoming contiguously and less coarsely punctate laterally; area behind and 
between ocelli weakly rugulose. (5) Gena with numerous moderately fine, 
contiguous carinae extending postero-dorsally from antero-ventral margin. 
(6) Malar area amber; very short. (7) Mandible yellow with apical % 
ferruginous. (8) Antenna with scape sometimes black, but usually yellow 
with brown or black on apical l / z of upper surface; pedicel brown below, 
brown-black above; flagellum brown to brown-black above, pale amber to 
brown below. Mesosoma: pubescence white, rarely tinged faintly with 
very pale yellow on mesoscutum. (9) Pronotum as in 6 . (10) Mesoscutum 
with medium-sized, moderately deep, contiguous punctures, becoming finely 
rugose anteriorly and moderately rugose antero-laterally. (11) Mesoscutel- 
lum with punctures like those of mesoscutum but sometimes less dense and 
subcontiguous laterally. (12) Metanotum moderately finely and shallowly 
rugose. (13) Mesepisternum moderately coarsely and deeply rugose. (14) 
Metepisternum rugose (like mesepisternum) to horizontally carinate with 
carinae anastomosing most often near posterior margin. (15) Propodeum 
with strong propodeal carina; propodeal shield very faintly rugose; dorsal 
area moderately to coarsely rugose; lateral area with shallow, horizontal, 
anastomosing rugae, very fine to moderately fine anteriorly, becoming much 
coarser and slightly deeper posteriorly. (16) Wing as in $ . (17) Tegula 
as in 6 but with metallic green tints on proximal margin. (18) Fore leg 
with coxa brown to brown-black, tinged with metallic green; trochanter 
sometimes black, but usually pale amber with yellow on distal l / 2 of anterior 
surface; femur entirely yellow to brown-black with small yellow spot dorsally 
at apex; tibia yellow to brown with yellow antero-dorsally on basal %; and 
tarsus yellow to brown. Pubescence white to amber. (19) Middle leg like 
fore leg but dark specimens with faint metallic tints on trochanter and with 
yellow area on tibia reduced to small spot dorsally at base; pubescence like 
that of fore leg but faintly fuliginous on darker specimens. (20) Hind leg 
with coxa bright metallic green dorsally and brown, strongly tinged with 
metallic green elsewhere; trochanter brown to brown-black; femur, tibia 
and tarsus pale amber to brown-black. Pubescence white to pale amber, be- 
coming fuliginous basally on dorsal area of tibia of darker specimens. 



Revision of the Bee Genus Agapostemon 503 

Metasoma: (21) Terga pale amber with small black spots laterally on 
terga 3-4 to entirely black with distal margins transparent (see preceding 
discussion of variation). Pubescence white, becoming pale amber around 
pygidium; bands of dense tomentum on basal halves of terga 2-5, and cen- 
trally on tergum 1 (often interrupted medially on tergum 1); short, ap- 
pressed, inconspicuous hairs distal to bands of tomentum; moderately long 
hairs anteriorly on tergum 1, ventro-laterally on terga 1-5, and dorsally on 
terga 5-6. (22) Sterna pale amber to black with scattered, long, white hairs 
on exposed areas. 

Agapostemon mexicanus n. sp. 

The name of this species refers to the country where it is found. 

The male holotype and the three male paratypes are labeled, "14 mi. 
S./El Arco Mine/L. Cal. VI-23-38//Michelbacher & Ross Collectors," The 
allotype is labeled, "Mesquital/L. Cal. VII-28-38//Michelbacher & Ross 
Collectors." All of these types are in the California Academy of Sciences, 
San Francisco. 

Distribution. In addition to the localities listed above I have seen 
specimens from: Santa Maria Valley (3 6 , 11 Aug.), Magdalena Bay 
(1 $ , 30 May), and 28 miles south of El Arco, Baja California, Mexico 
(1 9,3 July); 25 miles south of Navajoa, Sonora, Mexico (1 9 , 13 Sept.); 
17 miles south of Navajoa, Sonora, Mexico (1 $ , 18 Aug.). 

Diagnosis. The male may be distinguished from all other North Amer- 
ican species of Agapostemon by the comb of 14-20 large, evenly spaced sub- 
marginal bristles on metasomal sternum 4. The female may be distinguished 
from A. s plena 7 ens, A. texanus, A. angelic us, A. jemoratus, A. cocker elli and 
A. aenigma by its black metasoma; from A. melliventris, A. peninsularis, 
A. nasutus, A. leunculus and A. intermedins by the absence of yellow on its 
clypeus; from A. erebus and A. atrocaer ideas by the bright metallic green 
to blue-green of its head and mesosoma; and from A. tyleri by the lack of 
a smooth shiny area (with scattered punctures) centrally on its supraclypeal 
protuberance. 

Description 

male (Figs. 90-91, 156, 189, 204) 

General coloration of head and mesosoma bright metallic green to blue- 
green, metasoma banded with black and yellow. Head (Figs. 90-91): 
pubescence white, becoming yellowish on interocular area and vertex. (1) 
Labrum as in A. texanus but slightly more obtuse apically. (2) Clypeus 
as in A. melliventris but with medial enlargement of yellow band not quite 
as prominent. (3-7) Interocular area, vertex, gena, malar area and mandible 
as in A. melliventris. (8) Antenna brown-black with scape and pedicel 



504 The University Science Bulletin 

yellow below, flagellum pale amber below. Mesosoma: pubescence 
white, becoming yellowish on mesonotum and metanotum. (9-12) Prono- 
tum, mesoscutum, mesoscutellum and metanotum as in A. melliventris. 
(13) Mesepisternum more coarsely rugose than metanotum (especially 
anteriorly). (14-15) Metepisternum and propodenm as in A. melliventris 
but with finer sculpturing. (16) Wing as in A. melliventris but veins darker 
brown. (17) Tegula as in A. melliventris but with metallic green tints 
basally and with transparent areas very pale amber. (18-20) Fore, middle 
and hind legs as in A. peninsularis but with brown-black streak on tibiae 
posterior (Fig. 156). Metasoma: (21) Terga as in A. peninsularis but 
with pubescence shorter and less dense. (22) Sterna 2-4 yellow with brown 
submarginal band posteriorly, sterna 5-6 brown, sternum 1 with metallic 
green tints basally; pubescence long and pale on brown areas; sternum 4 
with low submarginal ridge posteriorly and row of 14-20 large, evenly spaced 
bristles just distal to submarginal ridge. (23) Genitalia (Figs. 189, 204) 
with medium-sized ventral lobe on gonocoxite fringed with hairs. 

female (Figs. 63-64) 

General coloration of head and mesosoma bright metallic green, meta- 
soma black. Head (Figs. 63-64): as in A. melliventris but lower margin 
of clypeus black instead of yellow and with pubescence becoming slightly 
fuliginous at vertex. Mesosoma: pubescence white, becoming fuliginous 
on mesonotum and metanotum. (9-15) Pronotum, mesoscutum, mesoscutel- 
lum, metanotum, mesepisternum, metepisternum and propodenm as in A. 
melliventris but with finer sculpturing. (16) Wing hyaline with veins 
brown, radius much darker than other veins. (17) Tegula as in A. melli- 
ventris. (18) Fore leg brown-black with faint metallic green on coxa and 
with yellow to pale amber anteriorly and dorsally on tibia and apically on 
femur. (19-20) Middle and hind legs brown to brown-black with hind 
coxa metallic green dorsally and tinged with metallic green anteriorly. 
Pubescence dark brown dorsally on tibiae. Metasoma: (21) Terga 
brown-black to black. Pubescence white becoming brown-black around 
pygidium; moderately long, becoming short and appressed dorsally; dense 
tomentum on basal third of terga 2-5, sparse tomentum on anterior % of 
tergum 1. (22) Sterna brown to brown-black with metallic green tints basally 
on sternum 1; pubescence white and moderately long. 

Agapostemon mourei n. sp. 

This species is named after Padre J. S. Moure who has contributed so 
much to knowledge of the South American bee fauna. 

The male holotype, the allotype and the paratypes (16 $ , 2 2 ) are all 
labeled, "Region Chapare/ Bolivia 400 M./ VIII 1950 Zischka." All of 



Revision of the Bee Genus Agapostemon 505 

these types are in the Snow Entomological Museum at the University of 
Kansas. 

Distribution. This species seems to be sympatric with A. boliviensis. 
In the Snow Entomological Museum I have seen additional specimens (14 

3,3 9) from the type locality and specimens labeled as follows: "BO- 
LIVIA-Beni, Rur-/ renabaque, 175 mts./ 5 October 1956/ (L. Pena)" (6 

(5); same as preceding but collected Oct. 17 (3 $); "BOLIVIA-Beni, 
Rey-/es. 10 December/ 1956 (L. Pena)" (2 $); "BOLIVIA- Chapare,/ 
Tropical Reg. 400/ mts. August 1951 (Zischka)" (3 $); same as preceding 
but 25 August 1950 (1 $ ) ; same as preceding but Oct. 1953 (5 6 ) ; "Cristal 
Mayu./ Chapare, Cochabamba Bol. 200 m./ VIII-4-49 L E Pena" (1 8 ); 
"Prov. del Sara/ Bolivia, 450 m./ J. Steinbech// Ace. 4548" (16); same 
as preceding but Ace. 5072 (1 6 ). In the U.S. National Museum I have 
seen specimens labeled: "nr. mouth/ Rio Mapiri/ Sept// Mulford/ Bio- 
Expl/ 1921-22" (35); "Covendo/ Boliv./ W M Mann// September// Mul- 
ford/ BioExpl/ 1921-22" (19); "Ixiamas/ Bolivia/ W M Mann// Dec// 
Mulford/ BioExpl/ 1921-22" (2$); "Rosario, Lake/ Rogagua, Boliv/ W M 
Mann, Oct / 28-Nov. 9, 1921// Mulford/ Biol. Expl.,/ 1921-1922" (2 $). 

Diagnosis. The male may be distinguished from most other species by 
the presence of a medial "button" at the base of its last visible (6th) sternum. 
This sternal "button" is narrowed basally unlike that of the closely related 
A. inca (Figs. 139-140), and the hind femur, tibia and basitarsus are less 
slender than those of A. inca (Figs. 167-168). Agapostemon mourei may be 
distinguished from the related A. lanosits, A. boliviensis and A. intermedins 
by the larger and differently shaped medial plate of its gonostylus (Figs. 
176,177A-B, 192). 

The female may be distinguished from most other species by the acute 
lateral angle and posterior lobe of its pronotum (Figs. 132-133) and by the 
dense woolly pubescence of its mesoscutum. It may be distinguished from A. 
lanosits by its bright yellow clypeal band and shallower genal striae; from 
A. boliviensis by its finer sculpturing and paler pubescence; from A. 
heteriirus by the yellow on its clypeus and by the coppery luster of its 
head and mesosoma. 

Variation. The general appearance of the males varies strikingly be- 
cause the black markings of the appendages and metasoma are frequently 
replaced by pale amber markings. 

Description 

male (Figs. 127-128, 132-133, 139, 167, 176, 217) 

General coloration of head and mesosoma bright metallic green to blue, 
metasoma with black and yellow or amber and yellow bands. Head (Figs. 
127-128) : with short creamy to yellow pubescence on interocular area and 



506 The University Science Bulletin 

vertex, becoming long and creamy to white on gena. (1) Labrum opaque 
lemon yellow with transparent pale amber margins, % as long as wide; 
anterolateral margins forming obtuse angle at apex; large transverse ridge 
on basal x / 3 without median depression. Punctures on crown of basal ridge 
shallow and separated by several times their diameters; submarginal punc- 
tures large and shallow at apex becoming smaller laterally and disappearing 
just anterior to lateral margins of basal ridge. Pubescence restricted to 
minute scattered hairs on crown of basal ridge and single submarginal row 
of simple hairs deflected distally from anterolateral margins. (2) Clypeus 
with large, shallow, sparsely scattered shallow punctures on lower (yellow) 
Vi-Vi) becoming more abundant, slightly smaller and contiguous on upper 
(green) x / 3 - x /i. (3) Interocular area with small, deep, contiguous punctures 
below ocelli, becoming medium-sized at level of antennal sockets and 
shallower, subcontiguous and smaller below antennae; supraclypeal pro- 
tuberance with medium contiguous moderately deep punctures. (4) Vertex 
with small, deep, contiguous punctures anteriorly and laterally, becoming 
shallower and slightly larger between ocelli; transversely carinate posteriorly. 
(5) Gena with parallel, subcontiguous, short, weak carinae extending dorso- 
lateral^ from antero-ventral margin. (6) Malar area yellow, short, and 
with minute hairs. (7) Mandible yellow with apical x /$- x / 3 amber. (8) 
Antenna dark brown to black, scape and pedicel yellow below. Mesosoma: 
pubescence short, pale fulvus to yellow on mesoscutum and mesoscutel- 
lum, becoming shorter and creamy- to white posteriorly and ventro- 
laterally. (9) Pronotttm (Figs. 132-133) with lateral angle and posterior 
lobe acutely pointed, lateral angle enlarged and projecting antero-laterally; 
with conspicuous sharp carina extending postero-ventrally from apex of 
lateral angle; with weak irregular rugae postero-laterally. Pubescence 
above lateral angle and posterior lobe pale fulvus to yellow, creamy to white 
below. (10) Mesoscutum with very large, conspicuous flange laterally on 
anterior margin extending antero-dorsally; punctures very fine, deep and 
contiguous. (11) Mesoscutellum with punctures slightly larger and shallower 
than those of mesoscutum and subcontiguous. (12) Metanotum deeply and 
moderately coarsely rugose. (13) Mesepisternum coarsely rugose anteriorly, 
becoming moderately rugose posteriorly. (14) Metepisternum moderately 
rugose with horizontal rugae slightly more prominent than other rugae. 
(15) Propodeum with propodeal carina nearly obliterated by extremely 
coarse and deep rugae posteriorly and dorsally; posterior and dorsal rugae 
becoming moderately fine, deep, contiguous punctures antero-laterally. (16) 
Wing sub-hyaline, tinted with amber on distal margins, veins and stigma 
brown. (17) Tegula pale transparent amber with anterior, submarginal, 
yellow crescent. (IS) Fore leg yellow with brown streak posteriorly on coxa, 
trochanter, femur and tibia. (19) Middle leg elongate, yellow with brown 



Revision of the Bee Genus Agapostemon 507 

markings on posterior of trochanter, tibia and sometimes femur. (20) Hind 
leg (Fig. 167) : femur sub-globose, tibia and basitarsus inflated. Coxa yellow 
below, dark brown with strong metallic tints above, trochanter dark brown, 
femur dark brown with yellow antero-dorsally, tibia dark brown sometimes 
streaked with yellow dorsally, and basitarsus varying from dirty yellow to 
dark brown. Metasoma: (21) Terga 2-6 very pale amber to dark brown- 
black on posterior 2 / 3 and yellow on anterior l / 3 (often partially concealed 
by overlap), tergum 1 with amber to brown-black band on anterior margin 
and brown to amber on anterior face; large, conspicuous, subcontiguous 
punctures anteriorly on tergum 1 becoming smaller posteriorly on tergum 
1 and progressively smaller on posterior terga. Pubescence sparse and yel- 
lowish. (22) Sterna dark brown and with yellow or amber markings on 
sterna 1-3; sternum 6 (Fig. 139) with proximo-medial projection, or "button," 
with each lateral flange slightly more than l / 2 width of base (as measured on 
transverse section through widest part of "button") ; pubescence moderately 
long, scattered, and pale fulvus to yellowish on exposed areas. (23) Genitalia 
(Figs. 176, 217). 

female (Figs. 125-126) 

General coloration dark copper with tints of metallic green. Head 
(Figs. 125-126) : with pubescence fulvus, becoming creamy on genal region. 
(1) Labrum dark amber to ferruginous. Basal ridge prominent, proximal 
face sloping gently toward proximal margin, distal face sharply declivous 
and delimited from basal area by carina; additional carina present between 
basal area and distal process; distal region narrow, rounded apically and 
with large central keel. Pubescence restricted to single submarginal row of 
coarse, stiff, amber bristles curving anteriorly at their apices and forming 
fimbria on antero-lateral margins of labrum. (2) Clypeus with shallow 
medial concavity; submarginal, yellow band broadly interrupted medially; 
punctures widely scattered on lower l / 2 , becoming sub-contiguous on upper 
1 /2- (3) Interocular area with fine, moderately deep, contiguous punctures 
below ocelli, becoming finely rugose laterally and below; supraclypeal pro- 
tuberance with moderate-size, subcontiguous punctures. (4) Vertex with 
fine, moderately deep, contiguous punctures anteriorly and laterally, and 
with fine, transverse striae posteriorly. (5) Gena with regular fine carinae 
extending postero-dorsally from antero-ventral margin. (6) Malar area 
amber to brown-black, very short, pubescence minute, white, appressed 
hairs. (7) Mandible yellow with apical ! /3 ferruginous. (8) Antenna dark 
brown to brown-black with nagellum pale yellowish amber below. Meso- 
soma: pubescence short, dense, woolly and pale ferruginous on meso- 
and metanotum, becoming sparse, moderately long and slightly paler below. 
(9) Pronotum like that of $ (Figs. 132-133) but with dense fulvus hair 
above lateral angle and posterior lobe. (10) Mesoscutum with small, deep, 



508 The University Science Bulletin 

contiguous punctures. (11) Mesoscntelhtm with punctures slightly smaller 
and shallower than those of mesoscutum. (12) Metanotum finely puncto- 
rugulose. (13) Mesepisternum with moderately coarse, shallow rugae an- 
teriorly, becoming slightly finer posteriorly. (14) Metepisternum shallowly 
rugose with horizontal rugae predominant. (15) Propodeum with prom- 
inent propodeal carina; propodeal shield coarsely rugose, with horizontal 
rugae predominant; dorsal area coarsely and deeply rugose; lateral area 
with moderately coarse, deep striae extending from propodeal carina to 
anterior margin. (16) Wing as in $ . (17) Tegula as in male, but slightly 
darker amber. (18) Fore leg brown with amber anteriorly on femur and 
tibia as well as dorsally and posteriorly; pubescence amber, becoming brown 
dorsally on tibia and tarsus. (19) Middle leg brown with amber anteriorly 
on tibia and femur and posteriorly on femur; pubescence pale amber, be- 
coming dark brown to brown-black dorsally on tibia and tarsus. (20) Hind 
leg brown to brown-black, pubescence pale amber, becoming dark brown to 
black antero-dorsally on tibia and basitarsus. Metasoma: (21) Terga 
finely and contiguously punctate anteriorly with punctures becoming smaller 
and shallower posteriorly; amber with yellow on basal ! /3 of terga 2-5 and 
brown apically on tergum 5; pubescence short and sparse, pale amber 
anteriorly and dark brown posteriorly. (22) Sterna pale amber with brown 
basally on sterna 1, 4-6; long, sparse, yellowish pubescence on exposed 
areas. 

Agapostemon nasutus Smith 

Agapostemon nasutus Smith 1853. Type $, British Museum (Natural History). 
Agapostemon peruvianus Cameron 1903 (new synonymy). Type $, British Museum (Natural 

History). 
Agapostemon nasutus gualanicus Cockerell 1912a (new synonymy). Type $, U.S. National 

Museum. 
Agapostemon purpureopictus Cockerell 1924 (new synonymy). Type 9, California Academy 

of Sciences, San Francisco. 
Agapostemon melanurus Cockerell 1949 (new synonymy). Type 9. U.S. National Museum. 

I have seen the types of all the above forms except A. nasutus and A. 
peruvianus, which were examined for me by Michener. There is no doubt 
that they are all the same species. A. nasutus gualanicus Cockerell is an 
ill-defined geographic variant which does not seem to warrant formal 
recognition as a subspecies (see discussion of variation). 

Distribution (Fig. 7). Agapostemon nasutus is the most common and 
widespread of the tropical Agapostemon. Absent from the central plateau of 
Mexico, it occurs along the eastern coast as far north as Brownsville, Texas, 
and along the western coast as far north as Hermosillo, Sonora, and El Arco, 
Baja California. It is widespread in southern Mexico and Central America 
but has not been found in the wet Atlantic lowlands of Nicaragua, Costa 
Rica and Panama. It has been collected along the coast of Venezuela as 



Revision of the Bee Genus Agapostemon 509 

far east as Trinidad, in the Cauca and Magdalena valleys of Colombia, and 
west of the Andes as far south as lea, Peru. 

It has been collected at or near sea level throughout its range and at 
elevations as high as 8,500 ft. (2,591 m) at Calderon, Ecuador; 1,200 m 
(3,963 ft.) at San Jose, Costa Rica; 5,000 ft. (1,524 m) at Guatemala City, 
Guatemala; and 6,400 ft. (1,951 m) at Tecamachalco, Puebla, Mexico. It 
has been collected in South America in every month but October, in Central 
America in every month but December and in Mexico north of the Isthmus 
of Tehuantepec in every month but January. Thus it seems that this species 
flies most, if not all, of the year wherever it occurs. My field observations 
in Costa Rica indicate that it is hard to find toward the end of the dry season, 
when flowers are also scarce. 

Diagnosis. Males are easily distinguished from all other species by the 
unique "hog-nosed" appearance of the broad, concave, ventral portion of the 
short clypeus (Figs. 85-87) and by sternum 6 which is broad with laterally 
directed disto-lateral tufts of branched hairs usually visible from above. 
Females, as well as males, may be distinguished from all other species with- 
out acutely pointed lateral pronotal angles by a yellow or creamy spot at the 
apex of the rounded posterior lobe of the pronotum (Fig. 134). 

Variation. The females normally are metallic green on the head and 
thorax but in the central highlands of Central America (e.g., Gualan, 
Guatemala and San Jose, Costa Rica) the green is replaced by blue. This 
is the form named A. nasutus gitalanicus by Cockerell. The same blue 
color is also found in many females from the southern limits of the range 
(e.g., Lima, Peru) and the northern limits of the range (e.g., Brownsville, 
Texas and Hermosillo, Sonora) where it may be very dark purple-blue. 

Normally black, the abdomen is pale amber in the majority of females 
from coastal Venezuela. I have seen females with black or intermediate 
abdomens from this same region although these constitute less than 10% 
of the total sample (Fig. 7). 

The amount of yellow on the males and females is quite variable and 
is slightly correlated with the coloration of the head and mesosoma of females. 
Blue females fly with males usually more extensively marked with black 
or brown than are males from regions where the metallic color of the 
females is green. 

Description 
male (Figs. 85-87, 153, 195, 221-223) 

General coloration of head and mesosoma metallic green to blue-green 
or bronze, metasoma yellow with black bands. Head (Figs. 85-87): (1) 
Labrum yellow, 40% as long as broad, flattened basal area slightly concave, 
20% as long as broad, distal area triangular, 85 % as long as basal width and 
separated from it by sharp step. Distal area shiny, impunctate; basal area 



510 The University Science Bulletin 

with scattered punctures most abundant distally. Distal margin with row 
of about 32 widely separated bristles, stout medially, finer laterally, re- 
mainder of distal area glabrous; basal area with scattered, prostrate hairs 
directed distally, longest and most abundant distally and forming fringe 
on distal margin of basal area. (2) Clypeus with ventral area smooth and 
shiny, dorsal area shiny and coarsely punctate; flattened ventral area glabrous, 
at right angle to upper area which is densely covered with white to yellowish 
hairs. (3) Interocular area with dense, fine punctures below vertex, rugose 
laterally, coarsely punctate or rugose above antennal sockets; pubescence 
dense, white to yellowish hairs nearly as long as scape and much shorter 
along inner margins of eyes. (4) Vertex with dense fine punctures below 
and in ocellar triangle; coarsely and shallowly rugose behind ocellar triangle. 
(5) Gena with sparse, weak carinulae extending postero-dorsally from antero- 
ventral margin; pubescence dense, white (never yellowish), medially about 
1.5 times as long as on interocular area. (6) Malar area yellow; pubescence 
sparse, white, tomentose. (7) Mandible yellow with apical third transparent 
ferruginous; nearly straight, not strongly curved as in other species. Pubes- 
cence on flat outer surface and posterior margin white to yellowish and 
directed apically; single longitudinal row of short simple bristles on inner 
surface. (8) Antenna (except scape and last two flagellomeres) dark brown 
above and light brown to yellow below; scape yellow with brown on apical 
V4- 1 /? above; last flagellomere brown with pale spot on upper side at apex; 
penultimate flagellomere brown, becoming yellow or light brown on basal 
V2 of underside. Mesosoma: metallic green, blue-green or bronze, with 
yellow or creamy spot on apex of posterior lobe of pronotum; moderately 
dense, white or creamy pubescence evenly distributed on all but anterior 
and lateral surfaces of pronotum. (9) Pronotum with lateral angle and 
posterior lobe rounded; 2-4 inconspicuous parallel carinae slanting rearwards 
and downwards near postero-lateral margin. Tomentum on lower lateral 
area white. (10) Mesoscutum densely covered with moderately deep, con- 
fluent punctures, fine posteriorly, becoming shallowly rugose on anterior 
l A- l A- (H) Mesoscntellum with fine, moderately deep punctures separated 
by about their own diameters. (12) Metanotum moderately and shallowly 
rugose medially, rugulose laterally. (13) Mesepistermtm coarsely and shal- 
lowly rugose anteriorly, becoming moderately rugose posteriorly. (14) 
Metepistemum with weak, parallel, interconnected, horizontal carinulae. 
(15) Propodeutn coarsely and shallowly rugose laterally, rugae slightly 
coarser and deeper dorso-medially; propodeal carina weak, evenly rounded 
dorsally; propodeal shield weakly and irregularly carinulate. (16) Wing 
pale, transparent brown, not darkened apically, radius much darker than 
other veins. (17) Tegula yellow with large transparent area in center of 
posterior half and transparent on posterior, distal and anterior margins. 



Revision of the Bee Genus Agapostemon 511 

(18) Fore leg yellow, sometimes with brown on posterior surface of coxa, 
trochanter and base of femur. Pubescence on coxa, trochanter and basal l / 2 
of femur sparse, long, white and branched; on tarsus, tibia and apex of 
femur sparse, short, simple, white to yellowish. (19) Middle leg yellow, coxa 
brown; trochanter and posterior area of femur sometimes brown; pubescence 
like that of fore leg. (20) Hind leg (Fig. 153) with coxa brown, becoming 
green dorsally; trochanter brown; brown streak on basal % of posterior 
surface of femur; tibia and tarsus yellow; on some specimens with brown 
on basal % of tibia and all but anterior of femur; pubescence similar to 
that of preceding legs except posterior surfaces of basitarsus and femur 
which are densely covered with short, stiff, simple, yellow hairs directed 
distally. Metasoma: (21) Terga 1-6 transparent apically, each with 
subapical brown or black band slightly broader than transparent area (may 
seem broader if abdomen is flexed and brown basal band of underlying 
tergum shows through overlapping transparent margin of preceding 
tergum), central yellow band almost twice as wide as subapical black band, 
and narrow brown or black basal band; tergum 1 with additional yellow 
basal region; tergum 7 brown or black; pygidial plate fawn to yellow. Pu- 
bescence fine and short dorsally, longer laterally, at base of tergum 1, and on 
posterior margins of terga 5-7; white except on dark portions of terga 5-7 
where yellow or brown; terga shiny with small, shallow punctures separated 
by about their own diameters. (22) Sterna 2-5 yellow, each with subapical 
brown band, apical margin transparent, narrow brown basal band often 
concealed by overlap of preceding sternum but sometimes so broad as to 
be confluent with subapical brown band; sternum 1 brown, tinged with 
green basally, paler subapically, distal and lateral margins transparent; 
sternum 6 yellow with basal brown band and transparent distal margin; 
yellow areas on abdominal sterna sometimes partially or completely replaced 
by brown; sternum 6 very broad with disto-lateral bulge and longitudinal 
medial ridge; sternum 1 with few scattered long white hairs and thin 
fringe of short hairs on distal margin; sterna 2-5 with scattered long yellow 
hairs subapically, distal margins fringed with short, prostrate hairs, basal % 
of each with minute, simple, white hairs; sternum 6 with numerous long 
yellowish hairs conspicuously branched, short medially but with long disto- 
lateral tufts directed postero-laterally and usually visible from above. (23) 
Genitalia (Figs. 195, 221-223) far more complex than those of other species, 
and relatively large. 

female (Figs. 41-42, 134) 

General coloration of head and thorax metallic green to blue. Metasoma 

black or pale amber, with white bands of fine pubescence. Head (Figs. 

41-42): pubescence short ( l / 2 J A as long as shortest distance between antennal 

sockets) and white or yellowish on face, twice as long on gena and pure 



512 The University Science Bulletin 

white. (1) Labrum dark amber to ferruginous. Basal ridge prominent, 
proximal face sloping gently toward proximal margin, distal face sharply 
declivous, and apex of ridge with deep punctures; distal portion narrow, 
abruptly rounded at apex; conspicuous median keel on antero-distal portion, 
continued basally almost to basal ridge; anterior margin of keel flattened 
and slightly flared laterally; single row of dark amber, wide, flat, stiff 
bristles curving anteriorly at their apices forming fimbria on lateral margins 
of distal portion of labrum. (2) Clypeus green above with yellow band on 
lower Yi-Yi bordered with black; yellow almost entirely replaced with dark 
brown or black on darker specimens; median subapical area slightly concave, 
yellow band with large scattered punctures, upper green area with numerous 
short horizontal carinulae or horizontally elongate punctures. (3) Inter- 
ocular area coarsely and shallowly rugose, supraclypeal protuberance with 
dense, long, horizontal carinulae. (4) Vertex densely and finely punctate 
between, anterior to, and lateral to ocelli; coarsely and shallowly rugose be- 
hind ocelli. (5) Gena with numerous weak, contiguous rugulae running 
postero-dorsally from antero-ventral margin. (6) Malar area brown and 
very short. (7) Mandible yellow, apical Y 3 ferruginous. (8) Antenna 
dark brown to black, slightly paler brown on underside of flagellum. 
Mesosoma: pubescence (except on pronotum) nearly twice as long as 
on face, moderately dense, white laterally and below, but may be yellowish 
or fuliginous on mesoscutum and mesoscutellum of blue specimens. (9) 
Pronotum (Fig. 134) with yellow to creamy spot at apex of posterior lobe; 
lateral angle and posterior lobe rounded as in $ \ faint punctures on dorsal 
portion and faint parallel striae running downward and rearward on lower 
lateral area; pubescence white, very fine, short, and appressed on lower 
lateral area. (10) Mesoscutum as in $ . (11) Mesoscutellum finely and 
densely punctate anteriorly, becoming coarsely punctate posteriorly; shiny, 
less densely punctate area lateral to center. (12) Metanotum rugulose. (13) 
Mesepisternum coarsely rugose anteriorly, somewhat less coarsely so pos- 
teriorly. (14) Metepisternum as in $ . (15) Propodeum coarsely rugose 
dorsally and postero-laterally, becoming finely striate antero-ventrally; 
propodeal carina prominent, strongly curved dorso-laterally, weakly convex 
dorsally; propodeal shield with irregularly anastomosing carinulae running 
laterally and slightly upward from mid-line. (16-17) Wing and tegula 
as in $ . (18) Fore leg brown to black with yellow on anterior apical % 
of femur and basal % of tibia, may be more extensive on pale specimens; 
pubescence of coxa, trochanter and femur white, becoming yellow on tibia 
and tarsus. (19) Middle leg similar in color and pubescence to fore leg but 
pubescence on dorsal surface of tibia and tarsus fuliginous. (20) Hind leg 
brown to black, pubescence similar to that of middle leg but even darker on 
dorsal surface of tibia and tarsus and with pale penicillus at apex of 



Revision of the Bee Genus Agapostemon 513 

basitarsus. Metasoma: (21) Terga usually black but often pale amber 
with black at apices of terga 4-6 in Colombian and Venezuelan specimens; 
anterior vertical portion of tergum 1 with sparse white hairs, upper x /z w i m 
thin white tomentum; basal halves of terga 2-4 with dense white tomentum; 
prepygidial fimbria and hairs adjacent to pygidium and on apical portion 
of tergum 4 brown to black. (22) Sterna brown to black, often pale amber in 
Colombian and Venezuelan specimens; long, scattered, whitish pubescence 
on exposed portions. 

Agapostemon obscuratus Cresson 

Agapostemon obscurata Cresson 1869. Type $ (not 9), Academy of Natural Sciences, 

Philadelphia. 
Agapostemon obscurants. Dalla Torre, 1896. 
Agapostemon obscuratus abjectus Cockerell 1917b (new synonymy). Type $ (not 9), U.S. 

National Museum. 

By peculiar coincidence the descriptions by both Cresson and Cockerell 
are headed by the symbol " 9 ", but both descriptions are of males. I have 
seen the types of A. obscuratus and A. obscuratus abjectus and both speci- 
mens are males. The color differences described by Cockerell were not 
visible to me. 

The specific rank of A. obscuratus is questionable, as I am unable to find 
any differences between specimens of A. viridulus and A. obscuratus other 
than the replacement of metallic coloration on the former by dark green- 
black on the latter. The sympatric distribution of A. obscuratus and A. 
viridulus and the absence of intermediate forms seems to refute any con- 
tention that they are merely subspecies. It is possible that they are a single 
species with two color phases. In the absence of biological data, I choose 
the simplest solution and regard A. obscuratus as specifically distinct from 
A. viridulus. 

Distribution. I have seen specimens from Habana, Cuba (Jan., 4 8 ; 
Feb., 2 $ ; April, 1 6 ; Sept., 1 8 ; Dec, 3 i ; 15 £ and 2 9 , no date). 

Diagnosis. Both sexes may be distinguished from all other West Indian 
species of Agapostemon by their total lack of metallic coloration. 

Description. As in A. viridulus but for the color differences mentioned 
above. 

Agapostemon ochromops n. sp. 

The name of this species refers to the pale yellowish eyes of both sexes. 

The male holotype is labeled, "Bahamas/Mariguana [Mayaguana?] /22- 
II//Greenway/coll." The female allotype and paratype are labeled, "Ba- 
hamas/Mariguana [Mayaguana?] /25-II // fls. Ctf/^7<?//Fairchild/coll." All 
of these types are in the Museum of Comparative Zoology, Harvard Univer- 
sity, Cambridge, Massachusetts. 

Distribution. Aside from the type series listed above, I have seen 



514 The University Science Bulletin 

specimens collected on the following Bahamian islands: Rum Cay (Feb.; 23 
$ , 3 9 ), Cat I. (March; 1 9 ), Concepsion [Conception?] I. (Feb.; 9 S , 
1 9 ) and New Providence I. (1 S ). 

Diagnosis. Both sexes may be distinguished from A. centratus, A. 
/{ohlielhis, A. obscurants, A. hispaniolicus, A. viridulus, A. swainsonae, 
A. cubensis, A. alayoi, A. insularis, A. jamaicensis, A. poeyi, A. viequesensis, 
A. columbi and A. aenigma by their pale milky eyes and creamy to yellow 
pigment on the underside of the pterostigma. The female is metallic green 
in contrast to the metallic dark blue of A. sapphirinus and the metallic dark 
purple of A. cyaneus. The male may be distinguished from A. cyaneus by 
the pale yellow to amber of the underside of its pedicel but is indistinguish- 
able from males of A. sapphirinus. 

Variation. The anterior surface of the first metasomal tergum of the 
allotype is honey-colored with an interrupted yellow band. On the other 
males the honey-color is replaced by light to dark brown. The brown on the 
legs is lightest on the allotype and darker on the other males. Metallic 
color varies in intensity and varies from blue-green to green both between 
and within island populations. The genal carinulae of females are finest 
and most numerous on the specimen from* Conception I., most coarse and 
fewest on specimens from Rum Cay, and intermediate on the holotype and 
paratype from Mariguana (Mayaguana?) I. 

This is the most heterogeneous of the Bahamian species and it may be a 
composite species. Only more collections from the Bahamas can resolve 
the problem. 

Description 
male (Fig. 225) 

General coloration of head and mesosoma bright metallic green to blue- 
green, metasoma brown to amber banded with yellow. Head: as in A. 
viequesensis but with eyes pale milky to yellowish, and ferruginous apical 
portion of mandible slightly reduced. Mesosoma: as in A. viequesensis 
but with slightly coarser sculpturing, with creamy to yellow pigment on 
pterostigma and on underside of larger wing veins. Metasoma: as in 
A. viequesensis but pygidium with apex more acute (Figs. 224-225), meta- 
somal terga 3-5 of the holotype and a specimen from Rum Cay with faint 
metallic tints and with pale bands slightly paler and dark bands slightly 
darker than in A. viequesensis, and genitalia with apex of distal stylus of 
gonostylus more spoon-shaped than in A. viequesensis (but not so much 
as in A. poeyi). 

FEMALE 

Like A. viequesensis but with eyes milky to yellowish; with creamy to 
yellow pigment on pterostigma and underside of larger wing veins; and with 



Revision of the Bee Genus Agapostemon 515 

sculpturing less coarse than in A. poeyi, but coarser than in A. vie que sen sis 
(closer to the latter), and somewhat more variable than in other West 
Indian species (even within island populations). 

Agapostemon peninsularis n. sp. 

The name of this species refers to the Baja California peninsula where 
it is found. 

The male holotype, the allotype and nine male paratypes are all labeled, 
"Mgdlena Bay/Lower Cal./May 30, 1925//H. H. Keifer/Collector" and 
are all in the California Academy of Sciences, San Francisco. 

Distribution. In addition to the type locality above I have seen specimens 
from the following localities: Ensenada (1 9 , 31 May; 1 $ , 24 June; 3 $ , 
30 Aug.), Camalu (1 2 , 23 June), 3 miles south of San Quintin (3 9 , 8-12 
March), 20 miles west of San Augustine (1 S , 24 Sept.), and 20 miles north 
of Mesquital (1 6 & 1 9, 27 Sept.), Baja California, Mexico; San Diego 
(IS) and La Jolla (1 S ,16 July), California. 

Diagnosis. The male may be distinguished from most species by its 
lack of a prominent hind femoral tooth and by the black streak extending 
from the base of the posterior surface of its hind femur to the apex of the 
posterior surface of its hind tibia; and from A. mexicanus by the lack of 
enlarged submarginal bristles medially on its 4th metasomal sternum (only 
two to four submarginal bristles are present on each side). The female 
may be distinguished from many species by its black metasoma, bright 
green head and thorax, transverse yellow clypeal band, and rounded lateral 
pronotal angle; from A. tyleri by the lack of a smooth shiny (with scattered 
punctures) area centrally on its supraclypeal protuberance; from A. nasutus 
by the absence of a creamy spot on the apex of its pronotal lobe; from A. 
leunculus by the white to creamy pubescence (brown to black in A. 
leunculiis) around the basitibial plate on its hind leg; from A. melliventris 
by the finely rugulose dorso-medial region of its propodeum (rugose in 
A. melliventris'). 

Description 

male (Figs. 88-89, 157, 190, 205) 

General coloration of head and mesosoma bright metallic green to blue- 
green, metasoma banded with black and yellow. Head (Figs. 88-89): 
pubescence white. (1) Labrum as in A. texanus but more obtuse apically. 
(2) Clypeus as in A. melliventris but with transverse yellow band slightly 
narrower and with medial enlargement only about l / 2 as great. (3-5) In- 
terocular area, vertex and gen a as in A. melliventris but with sculpturing 
slightly coarser. (6-7) Malar area and mandible as in A. melliventris. (8) 



516 The University Science Bulletin 

Antenna with scape and pedicel brown-black above and yellow below to 
entirely brown-black; flagellum brown-black above, becoming pale amber to 
yellow below. Mesosoma: pubescence white. (9-12) Pronotum, meso- 
scutum, mesoscutellum and metanotum as in A. melliventris. (13) Mesepi- 
sternum more coarsely rugose than metanotum (especially anteriorly). (14) 
Metepisternum with interconnected horizontal carinulae, becoming rugose 
posteriorly in some specimens. (15) Propodeum with propodeal carina 
usually absent but may be visible as very weak carina dorsally; propodeal 
shield very coarsely and shallowly rugose; dorsal area moderately rugose 
laterally, becoming very finely and shallowly rugose medially; lateral area 
coarsely and shallowly rugose posteriorly, becoming moderately finely rugose 
anteriorly. (16) Wing as in A. melliventris but with veins darker brown. 
(17) Tegula pale transparent amber with submarginal yellow band anteriorly 
and posteriorly, and with strong metallic tints basally. (18) Fore leg with 
coxa metallic green; trochanter brown-black with metallic green tints basally 
and apical yellow spot anteriorly; femur yellow with broad brown-black 
streak postero-dorsally extending from base almost to apex; tibia yellow 
with brown-black streak ventrally; tarsus yellow. (19) Middle leg as fore 
leg but with brown-black ventral streak on tibia extending to posterior sur- 
face. (20) Hind leg (Fig. 157) as fore leg but with yellow spot on trochanter 
ventral; with brown to brown-black streak on femur posterior, extending 
to apex, and extending dorsally at apex; tibia with brown to brown-black 
streak posterior and extending dorsally at base and apex. Femur without 
distinct tooth and basitarsus without apical groove or basal ridge. Meta- 
soma: (21) Terga 2-6 black with yellow band on basal l / z (usually 
partially obscured by overlapping terga posteriorly) ; tergum 1 black with 
yellow band centrally; terga 1-6 with posterior and ventro-lateral margins 
transparent; tergum 7 with pygidium dark amber to brown-black. Pubes- 
cence white, dark brown on postero-dorsal Y 2 of terga 1-4; moderately long, 
becoming short and appressed dorsally on terga 1-4. (22) Sterna 2-5 yellow 
with brown-black bands basally and anteriorly (these bands very narrow 
on sternum 2, becoming progressively broader on posterior sterna until 
yellow nearly absent on sternum 5), sternum 1 with strong metallic tints, 
sternum 6 brown-black with small medial ridge in shallow depression on 
apical Y 2 . Pubescence long and pale on exposed areas, sternum 4 with 2-4 
large hooked bristles at each postero-lateral margin. (21) Genitalia (Figs. 
190, 205) with medium-sized ventral lobe on gonocoxite fringed with hairs. 

female (Figs. 65-66) 

General coloration of head and mesosoma bright metallic green, meta- 
soma black. Head (Figs. 65-66): as in A. melliventris but with transverse 
band on clypeus narrower and bordered with brown-black, and with scape 



Revision of the Bee Genus Agapostemon 517 

entirely brown-black. Mesosoma: pubescence white, becoming creamy 
on mesonotum and pronotum. (9-14) Pronotum, mesoscutum, mesoscutel- 
lum, metanotum, mesepisternum and metepisternum as in A. melliventris. 
(15) Propodeum with weak propodeal carina; propodeal shield with large 
but very shallow rugae; dorsal area finely rugose laterally, becoming very 
finely rugulose medially; lateral area with shallow, anastomosing horizontal 
carinulae anteriorly, becoming shallowly rugose posteriorly. (16) Wing 
as in 6 . (17) Tegula as in A. melliventris. (18-19) Pore and middle legs 
with coxa dark brown tinged with metallic green; trochanter dark brown; 
femur dark brown with yellow apically; tibia yellow with dark brown 
streak ventrally; tarsus yellow. (20) Hind leg brown, coxa metallic green 
dorsally. Metasoma: pubescence white. (21) Terga brown-black to 
black. Pubescence moderately long, becoming short and appressed dorsally 
on terga 1-4; dense tomentum on basal ! / 2 of terga 2-5 and on all but posterior 
! / 3 of tergum 1. (22) Sterna brown-black to black; moderately long pubes- 
cence on exposed areas. 

Agapostemon poeyi (Lucas) 

Andrena poeyi Lucas 1856. Type 2, Gribodo collection, Museo Civico di Storia Naturale, 

Genoa. 
Agapostemon jestivus Cresson 1865 (new synonymy). Type $, Academy of Natural Sciences, 

Philadelphia. 
Halictus (Agapostemon) chalets Vachal 1903 (new synonymy). Type $, Naturhistorisches 

Museum, Vienna. 

The "type" of Andrena poeyi has been examined by C. D. Michener. 
He states that it is labeled "Andrena poeyi Guerin Cuba Type" but notes 
that the word "type" is not in Guerin's hand. While this specimen may 
not be the type, I am regarding it as such until such time as a more likely 
candidate is found. The description coupled with the type locality leave 
little doubt as to the identity of this species. I have seen the type of A. 
jestivus and it is obviously the male of A. poeyi. I have seen the type of 
H. chalets and it is unquestionably conspecific with A. poeyi. 

Distribution. I have seen about 70 specimens from Cuba; males have 
been collected in all months but January, June, October and December and 
females have been collected in January, April, June, September and Novem- 
ber. I have also seen specimens from Guaiameti, Santo Domingo, Hispaniola 
(IS, July); Arthurs Town, Cat I., Bahamas (19, July-Aug.); Mangrove 
Cay [part of Andros I.], Bahamas (19,1 Aug.); Nassau, New Providence 
I, Bahamas (5 9 , 12 Aug.; 1 9 , Nov.; 1 6 , 16 Dec; 7 6 & 3 9 ). 

I have also seen three males of this species labeled "Mexico-Veracruz,/ 
Nogales/April 1956/ (N. L. H. Krauss)." In his revision of the banded sub- 
genera of Nomia, D. W. Ribble (1965) noted specimens of the Cuban Nomia 
robinsoni with these label data and questioned the correctness of the labels. 



518 The University Science Bulletin 

In the face of this additional evidence it seems almost certain that the 
labels are incorrect. 

Diagnosis. Both sexes may be distinguished from those of A. cyanens, 
A. sapphirinus and A. ochromops by their lack of milky to yellowish eyes 
or pterostigma. The male may be distinguished from A. columbi and A. 
viequesensis by its strong metallic green to blue tints medially on metasomal 
tergum 3; from A. insularis by its metallic blue to green tints postero- 
laterally on tergum 1; and from A. jamaicensis by its narrower hind femora. 
The female may be distinguished from A. centratus, A. alayoi, A. obscuratus, 
A. hispamohcus, A. viridulus and A. swainsonae by its bright metallic green 
metasomal terga; from A. viequesensis by the basal green spot on its man- 
dible; from A. columbi and A. aenigma by its conspicuous metallic green 
tints medially on metasomal sterna 3-4; from A. insularis by its rugose area 
adjacent to the parapsidal line. 

Description 

male (Figs. 144, 186, 210) 

General coloration of head and mesosoma bright metallic green to blue- 
green, metasoma dark brown with yellow bands. Head: color and shape 
as in A. viequesensis; sculpturing very slightly coarser than in A. vieque- 
sensis; pubescence of clypeus, interocular area and vertex yellowish to pale 
amber, that of gena white. Mesosoma: pubescence as in A. viequesensis. 
(9) Pronotum with lateral angle and posterior lobe more angular than in 
A. viequesensis; carinulae on postero-lateral area larger and fewer in number 
than in A. viequesensis. (10-12) Mesoscutum, mesoscutellum and metano- 
tum as in A. viequesensis but with very slightly coarser sculpturing. (13-15) 
Mesepisternum, metepi sternum and propodeum as in A. viequesensis but 
with much coarser sculpturing. (16) Wing as in A. viequesensis but with 
veins darker. (17) Tegula with dark amber to brown almost obscuring yellow 
band, metallic green tints antero-basally slightly more extensive than in 
A. viequesensis. (18-19) Fore and middle legs with coxae metallic green 
(only weakly tinted on middle leg), trochanters brown with amber to 
yellow on anterior surface, femora yellow with apical, brown postero-dorsal 
spot, tibiae yellow with brown postero-dorsally, and tarsi yellow to pale 
amber (some specimens from New Providence I. with brown areas on fore 
leg reduced) ; pubescence slightly darker than that of A. viequesensis. (20) 
Hind leg (Fig. 144) as in A. viequesensis but with brown areas much 
darker, with femur, tibia and basitarsus less swollen, and with ridge and 
apical groove on basitarsus narrower and less prominent. Metasoma: 
(21) Terga as in A. viequesensis but with brown areas darker, with apex of 
pygidium nearly as acute as that of A. ochromops (Figs. 224-225), and with 
metallic green replacing brown completely on terga 4-6 and partially on 2-3. 



Revision of the Bee Genus Agapostemon 519 

(22) Sterna as in A. vie que sen sis but slightly darker and with small metallic 
medial green spot on transverse submarginal ridge of sternum 4. (23) 
Genitalia (Figs. 186, 210) with apex of apical stylus on gonostylus enlarged 
like bowl of spoon. Ventral lobe large, with fringe of hairs on lower margin. 

female (Figs. 8, 10, 12, 14, 16) 

General coloration of head, mesosoma and metasoma bright green to 
blue-green. Head: pubescence as in A. viequesensis. (1) Labrum as in 
A. hispaniolicus. (2) Clypeus with broad, deep punctures on brown-black 
portion becoming deeply rugose on upper l / 2 . (3) Interocular area deeply 
rugose, with rugae largest just above antennal sockets; supraclypeal pro- 
tuberance rugose. (4) Vertex finely rugose anteriorly, becoming coarsely 
rugose laterally and posteriorly. (5) Gena with less than ten extremely 
large rugae extending postero-dorsally from antero-ventral margin (Fig. 16). 
(6) Malar area as in A. viequesensis. (7) Mandible as in A. viequesensis 
but with basal area darkened and with metallic basal spot. (8) Antenna 
brown-black, slightly paler on underside of flagellum, apex of apical flagel- 
lomere sometimes conspicuously paler than other flagellomeres. Meso- 
soma (Figs. 8, 10, 12, 14): with pubescence as in A. viequesensis. (9) 
Pronotum with lateral angle and posterior lobe slightly angular; dorsal por- 
tions of lateral angle and posterior lobe rugose and separated from lower 
portion by distinct carina extending from lateral angle to posterior lobe; 
postero-lateral area with 3-6 strong horizontal carinae. (10) Mesoscutum 
very coarsely rugose laterally and anteriorly, becoming coarsely and con- 
tiguously punctate medially and posteriorly. (11-12) Mesoscutellum and 
metanotum as in A. viequesensis. (13) Mesepistemum evenly, extremely 
deeply and coarsely rugose. (14) Metepisternum with extremely coarse 
horizontal carinae. (15) Propodeum with very strong propodeal carina; 
propodeal shield with strong carinae extending dorso-laterally from medial 
groove; dorsal area with extremely coarse carinae extending postero-laterally 
from antero-dorsal margin; lateral area with coarse, anastomosing carinae 
extending from metepisternum to lateral portion of propodeal carina. (16) 
Wing as in A. viequesensis but with veins darker. (17) Tegula as in A. 
viequesensis but with amber replaced by dark brown, metallic tints slightly 
more extensive, and yellow band largely or wholly obscured. (18-20) Fore, 
middle and hind legs dark brown with coxae wholly to partially covered 
with bright metallic green (least on middle coxa); pubescence as in A. 
viequesensis but darker (almost black) dorsally on middle tibia, hind tibia 
and apex of hind femur. Metasoma: (21) Terga as in A. viequesensis 
but with central brown bands absent or very faint. (22) Sterna dark brown 
with strong metallic tints on all or most of exposed portions; pubescence as 
in A. viequesensis. 



520 The University Science Bulletin 

Agapostemon radiatus (Say) 

Halictus radiatus Say 1837. Type probably destroyed. 

Agapostemon pulchra Smith 1853. Type $ , British Museum (Natural History). 

Agapostemon sitlcatulits Cockerell 1909. Type $ , U.S. National Museum. 

The type of Halictus radiatus has probably been destroyed, but the 
description leaves no doubt of its identity. Michener has examined the 
type of Agapostemon pulchra (which is not labeled "California" as reported 
by Sandhouse, 1936) and I have examined the type of Agapostemon sul- 
catulus. 

Distribution (Fig. 18). Specimens have been collected as far north as 
Treesbank, Manitoba; as far south as Orange Co. in central Florida; as 
far east as Cape Cod, Massachusetts; and as far west as Sioux Co., Nebraska. 
In the northern part of the range (Wisconsin) females have been collected 
from April through October and males from July through October. In the 
central part of its range (Kansas)- females have been collected from April 
through November and males in April and July through November. In the 
southern part of its range females have been collected from April through 
September and males from June through August and in October. Most 
abundant at elevations of less than 2,000 ft. (610 m), A. radiatus has been 
collected as high as 4,100 ft. (1,250 m) in the Smoky Mountains of North 
Carolina. 

Diagnosis. The male may be distinguished from many species by the 
yellow maculations on its 6th metasomal sternum and by the tooth on its 
hind femur; from A. mellwentris by the brown-black on the anterior surface 
of its first metasomal tergum; from A. texanus and A. angelicus by its lack 
of conspicuous metallic tints on metasomal tergum 4, and by its lack of a 
broad, subapical, transverse ridge on metasomal sternum 4; and from males 
of A. splendens, A. cocherelli and A. femoratus by its unmodified hind 
basitarsus. The female may be distinguished from many species by its 
metallic green metasomal terga; from A. texanus and A. angelicus by its 
coarsely punctate or rugose mesoscutum without punctures of two distinct 
sizes; from A. splendens by its smaller size, coarse parallel carinae dorso- 
medially on its propodeum, its bright yellow mandibles (ferruginous 
apically), and its very pale amber to clear wings; from A. femoratus by its 
mesoscutum (coarsely punctate with rugae anteriorly and laterally in A. 
radiatus, almost entirely coarsely rugose in A. femoratus). The female can- 
not be morphologically differentiated from females of A. cocherelli but may 
be distinguished by its eastern distribution. 

Description 

male (Figs. 105-106, 161) 

General coloration of head and mesosoma bright metallic green to 
blue-green, metasoma with black and yellow bands. Head (Figs. 105-106) : 



Revision of the Bee Genus Agapostemon 521 

pubescence white, pale yellowish on interocular area, becoming pale amber 
on vertex. (1) Labrum as in A. texanus. (2) Clypeits with large scattered 
punctures below, becoming weakly rugose above. (3) Interocular area with 
very deep, fine, contiguous punctures below vertex, becoming weakly and 
finely rugose at level of and below antennal sockets; supraclypeal pro- 
tuberance rugose. (4) Vertex finely and deeply punctate below ocelli and 
laterally, becoming deeply rugulose behind and between ocelli. (5) Gena 
with short, weak, parallel, well separated carinae extending postero-dorsally 
from antero-ventral margin. (6) Malar area bright yellow; very short. (7) 
Mandible bright yellow with apical x /^- x /i dark ferruginous. (8) Antenna 
brown-black above with apical l / 2 of apical flagellomere amber; underside 
of flagellum amber to pale amber; underside of scape and pedicel bright 
yellow (underside of pedicel sometimes brown). Mesosoma: pubescence 
white, becoming yellowish to fuliginous on mesonotum and metanotum. 

(9) Pronotum with distinctly, but not acutely, angular lateral angle and 
posterior lobe; distinct carina extending postero-ventrally from lateral angle; 
distinct vertical carina laterally; and numerous weak horizontal carinulae. 

(10) Mesoscutitm with fine, deep, contiguous punctures becoming rugose 
antero-laterally. (11) Mesoscutelhtm with fine, deep, contiguous punctures 
medially, becoming scattered in shiny lateral area. (12) Metanotum finely 
but deeply rugulose. (13) Mesepi sternum moderately deeply and moderately 
coarsely rugose anteriorly, becoming slightly less deeply and coarsely rugose 
posteriorly. (14) Metepisternum moderately deeply and moderately coarsely 
rugose. (15) Propodeum with strong propodeal carina; propodeal shield 
moderately deeply and moderately coarsely rugose; dorsal area with short, 
coarse, parallel carinae extending postero-laterally from anterior margin, 
becoming moderately deeply and moderately coarsely rugose posteriorly 
(entire dorsal area rugose on some specimens) ; coarsely rugose postero- 
laterally, becoming punctate or finely rugulose antero-laterally. (16) Wing 
very pale brown, slightly darker on distal margin; veins and pterostigma 
pale brown, radial vein very dark brown. (17) Tegula amber with metallic 
tints antero-basally, with yellow on posterior margin, and with yellow sub- 
marginal band anteriorly. (18) Fore leg yellow; coxa metallic green and 
with small brown stripe posteriorly on tibia; pubescence yellowish to amber. 
(19) Middle leg with coloration and pubescence like fore leg with small, 
brown, subapical spot dorsally on femur and with brown posterior stripe 
on tibia larger. (20) Hind leg (Fig. 161) with coxa metallic green; tro- 
chanter brown to amber with yellow on apical l / 2 of lower side (extending 
to base on some specimens) ; femur yellow with brown spot dorsally at 
apex and sometimes posteriorly at base (much smaller than apical spot); 
tibia yellow with small brown to amber spot apically on anterior surface, 
with brown stripe antero-dorsally often connected with brown basal spot; 



522 The University Science Bulletin 

tarsus yellow, basitarsus lacking basal ridge and apical groove. Pubescence 
yellowish to amber. Metasoma: (21) Terga black with yellow bands 
on basal halves of terga 2-6 and centrally on tergum 1 ; ventro-lateral margins 
of terga 1-6 transparent amber; and sometimes with faint metallic tints 
anteriorly on tergum 1 and postero-laterally on tergum 3. Pubescence dor- 
sally on terga 1-4 very short and inconspicuous, pale on yellow bands, dark 
on brown bands; moderately long, yellowish pubescence anteriorly on tergum 
1 and laterally on terga 1-4; moderately long and dark brown dorsally and 
laterally on terga 5-7. (22) Sterna yellow with narrow brown band on distal 
margin of sternum 2 pale and slightly broadened medially, becoming pro- 
gressively darker and larger on sterna 3-5; sternum 6 with brown medial 
stripe varying from very narrow to l / 2 width of sternum; sternum 1 metallic 
green basally. Broad, low, transverse, subapical (not apical as in A. texantis 
and A. angelicas) ridge on sternum 4, most prominent laterally, and often 
with faint metallic tints medially; moderately long scattered hairs on ex- 
posed areas, 2-4 large stout bristles disto-laterally on sternum 4. (23) 
Genitalia indistinguishable from those of A. femoratus. 

female (Figs. 55-56) 

General coloration of head, mesosoma and metasoma bright metallic 
green to blue-green. Head (Figs. 55-56): pubescence white, becoming 
pale yellowish amber on vertex and, commonly, on interocular area. (1) 
Labrum as in A. texanns. (2) Clypeus with punctures deep, scattered below, 
becoming subcontiguous and interspersed with weak horizontal rugae above. 
(3) Interocular area with coarse, deep, contiguous punctures dorso-medially, 
becoming moderately rugose laterally and ventrally; supraclypeal pro- 
tuberance sparsely punctate and with weak horizontal carinae. (4) Vertex 
with coarse, deep, contiguous punctures anterior to ocelli, becoming more 
shallowly punctate or finely rugulose laterally; shallowly and finely rugose 
posteriorly and between ocelli. (5) Gena with moderate to coarse, contiguous, 
parallel carinae extending postero-dorsally from antero-ventral margin. (6) 
Malar area dark amber to dark brown; very short, nearly absent. (7) Man- 
dible yellow with apical Yi-Vi dark ferruginous. (8) Antenna brown-black 
with underside of flagellum very slightly paler than upperside. Meso- 
soma: pubescence white, becoming slightly yellowish to fuliginous on 
mesonotum and metanotum. (9) Pronotum as in $ . (10) Mesoscutum with 
deep, contiguous, moderately large punctures becoming rugose anteriorly 
and antero-laterally. (11) Mesoscutellum with small subcontiguous punc- 
tures medially, becoming larger and scattered on shiny lateral area. (12) 
Metanotum finely and shallowly rugulose. (13) Mesepisternum deeply and 
coarsely rugose anteriorly, becoming slightly less coarsely and less deeply 
rugose posteriorly. (14) Metepi sternum with even, parallel, horizontal 



Revision of the Bee Genus Agapostemon 523 

carinae anastomosing posteriorly (on some specimens becoming rugose 
posteriorly). (15) Propodeum with very strong and prominent propodeal 
carina; propodeal shield weakly rugose (rugae often absent on upper 
margin) ; dorsal surface with very coarse, irregular and anastomosing carinae 
extending posteriorly from anterior margin; postero-lateral margin with 
moderately coarse horizontal carinae, becoming much finer and weaker 
antero-laterally. (16) Wing as in S . (17) Tegirfa as in $ but slightly 
darker. (18-19) Fore and middle legs dark brown to brown-black with 
coxae metallic green and with antero-dorsal pale amber to yellowish areas 
apically on femora and basally on tibiae; pubescence white to pale fuliginous 
or amber. (20) Hind leg dark brown to brown-black with metallic tints 
dorsally on coxa; pubescence creamy, becoming amber posteriorly on tarsus 
and dark brown to brown-black dorsally on tibia. Metasoma: (21) 
Terga with very fine subcontiguous punctures on metallic regions; pubes- 
cence white, becoming brown-black on terga 5-6; terga 2-4 with narrow 
band of tomentum basally, otherwise largely devoid of hairs; moderately 
long hairs anteriorly on tergum 1, laterally on terga 1-5, and dorsally on 
terga 5-6. (22) Sterna brown to brown-black with metallic tints basally on 
sternum 1 and some specimens with very weak metallic tints on sternum 4; 
exposed areas with long, white, scattered hairs. 

Agapostemon rhopalocera Smith 

Agapostemon rhopalocera Smith 1853. Type $, British Museum (Natural History). 

The type of this rare and peculiar species was examined at the British 
Museum by C. D. Michener. I know of only two additional specimens 
(males), both loaned to me by the British Museum. 

Distribution. The holotype is from "Mexico." The other two males 
were collected in May, 1913-14 at Atoyac, Vera Cruz, Mexico. This town 
lies just 3.5 km ESE of Pico de Orizaba at an elevation of about 600 m 
(1,968 ft.). As well as this area has been collected, it is surprising that no 
more specimens have been collected in the past 50 years. Perhaps the habitat 
of this species has been destroyed by human activities. 

Diagnosis. The male may easily be distinguished from all other members 
of Agapostemon by its long filamentous flagellum with the apical flagello- 
mere flattened (Fig. 131). 

Description 

male (Figs. 77-78, 131, 154, 207) 

General coloration of head and thorax metallic green, abdomen with 
black and yellow bands. Head (Figs. 77-78): (1) Labrmn as in A. texanus 
but slightly more acute apically. (2) Clypeus with punctation as in A. 
leunculus, pubescence white to yellowish. (3) Interocular area above an- 



524 The University Science Bulletin 

tennal sockets very coarsely rugose or with dense contiguous punctures, 
coarsely rugose at level of antennal sockets, becoming moderately rugose 
below antennal sockets; pubescence white to yellowish. (4) Vertex with 
sculpturing as in A. leunculus; pubescence white to yellowish. (5) Genu 
with sculpturing as in A. leunculus; pubescence white to yellowish as in 
A. leunculus but denser and medially more coarsely branched and bent 
medially so as to cover the proboscidial fossa. (6-7) Malar area and mandible 
as in A. texanus. (8) Antenna (Fig. 131) with apical flagellomere and distal 
% of subapical flagellomere dark brown to black, remainder of flagellum 
and pedicel brown above and yellow-amber below, scape amber with dark 
streak on distal % of upper surface. Mesosoma: pubescence moderately 
long, evenly distributed (except pronotum), and whitish to amber as in A. 
leunculus. (9) Pronotum as in A. leunculus. (10) Mesoscutum finely and 
contiguously punctate, sometimes finely rugose anteriorly and laterally. 

(11) Mesoscutellum with punctures separated by about l / 2 their diameters, 
slightly coarser than those of mesoscutum and becoming rugulose laterally. 

(12) Metanotum finely rugulose medially becoming shallowly rugose 
laterally. (13) Mesepisternum as in A. nasutus. (14) Metepisternum rugose 
with few weak horizontal carinulae. (15) Propodeum as in A. leunculus 
but with more prominent propodeal carina. (16) Wing as in A. leunculus. 
(17) Tegula as in A. nasutus but with reduced yellow area. (18) Fore leg 
as in A. leunculus but more slender; pubescence white to amber. (19) 
Middle leg very slender; coxa dark brown to black, remainder of leg yellow 
with brown-black on all but anterior surface of trochanter and with brown- 
black stripe on basal Vz-% of posterior surface of tibia; pubescence white to 
amber. (20) Hind leg (Fig. 154) very slender, coxa metallic green, trochanter 
brown-black, remaining segments yellow with brown streak on posterior of 
femur and brown at base of tibia; pubescence white to amber. Meta- 
soma: (21-22) Terga and sterna as in A. leunculus. (23) Genitalia (Fig. 
207) very similar to A. leunculus but penis valves broader and flatter, medial 
plate of gonostylus longer and narrower; and ventral lobe smaller and not 
notched. 

Agapostemon sapphirinus n. sp. 

The name of this species refers to the sapphire blue color of the female. 

I have seen 15 specimens (14 $ , 1 $ ) of this species. The female holo- 
type, the allotype and the 13 male paratypes are all labeled, "Bahamas/ 
Clarencetown/Long Isl. Febr//Utowana/Exp. 1934." All of the types are 
at the Museum of Comparative Zoology, Harvard University, Cambridge, 
Massachusetts. 

Diagnosis. Both males and females may be distinguished from all other 
West Indian species, except A. cyaneus and A. oc/iromops, by their milky to 
yellowish eyes and pterostigma. The male may be distinguished from 



Revision of the Bee Genus Agapostemon 525 

A. cyaneus by the pale amber to yellow on the underside of its pedicel, but 
cannot be reliably distinguished from A. ochromops. The female may be 
distinguished from A. ochromops by its dark purple metallic color and 
from A. cyaneus by its very coarsely carinate gena. 

Description 
male 

General coloration of head and mesosoma bright metallic green to dark 
metallic green to blue, metasoma with black and yellow bands. Head: 
as in A. viequesensis but with eyes cream-colored, underside of flagellum 
yellower, pubescence of interocular area slightly yellow, and sculpturing 
very slightly coarser. Mesosoma: pubescence as in A. viequesensis but 
slightly yellowish laterally and slightly darker on mesonotum and metano- 
tum. (9) Pronotum with lateral angle and posterior lobe more angular than 
in A. viequesensis. Carinulae on postero-lateral area larger and fewer in 
number than in A. viequesensis. (10-12) Mesoscutum , mesoscuteUum and 
metanotum as in A. viequesensis but with very slightly coarser sculpturing. 
(13-15) Mesepisternum , metepisternum and propodeum as in A. viequesensis 
but with much coarser sculpturing. (16) Wing as in A. viequesensis but 
with yellow pigment on pterostigma and underside of larger veins. (17) 
Tegula as in A. viequesensis but with metallic tints less extensive. (18-20) 
Fore, middle and hind legs as in A. viequesensis. Metasoma: (21) 
Terga as in A. viequesensis but with metallic tints on some specimens (never 
as extensive as in A. poeyi), with dark bands almost black, and with yellow 
bands paler. Apex of pygidium nearly as acute as in A. ochromops. (22) 
Sterna as in A. viequesensis but often yellower basally and darker along 
apical margin. (23) Genitalia as in A. poeyi (cf. Figs. 186, 210). 

FEMALE 

Female as in A. poeyi but deep metallic blue, with metallic tints on sterna 
less intense and less extensive, with indistinct brown to ferruginous bands 
on terga 1-4, with creamy pigment on pterostigma and underside of larger 
wing veins, and with eyes pale. 

Agapostemon semimelleus Cockerell 

Agapostemon semimelleus Cockerell 1900. Lectotype $ , Carnegie Museum, Pittsburgh. 
Halictus (Agapostemon) andensis Vachal 1903 (new synonymy). Type 9, Museum National 

D'Histoire Naturelle, Paris. 
? Agapostemon castanetis Schrottky l ( M)2b. Type $ ? Departamento de Zoologia, Sao Paulo? 

Schrottky (1909b) tentatively synonymized the female of this species with A. semimelleus. 

Not having seen a type I cannot dispute his judgement. 
? Agapostemon coryliventris Holmberg 1903. Type $. 
? Agapostemon multicolor Holmberg 1903. Type $. According to J. S. Moure (pers. comm.) 

the types of most Holmberg species have been destroyed by dermestids. Therefore I am 

following Schrottky (1909b) in his synonymy of A. coryliventris and A. multicolor with 



526 The University Science Bulletin 

A. semimelleus. Although these nominal species may be synonyms of A. chapadensis they 
would still be junior synonyms. 

I have seen the syntypes of A. semimelleus and have labeled the male as 
lectoholotype and the female as lectoallotype. Both specimens are from 
Chapada. I selected the male as lectoholotype because I am unable to dis- 
tinguish the females of A. semimelleus and A. chapadensis, although the 
female lectoallotypes are easily distinguished. 

Distribution. This species is widespread in the campos, or tropical grass- 
lands, of southern Brasil, of Paraguay and of northeastern Bolivia. It has 
also been collected at the heads of the Maraiion and Ucayali Valleys in Peru, 
the heads of the Putamayo and Cauca Valleys in Colombia, and along the 
eastern slopes of the Andes in Bolivia and northern Argentina. It ranges 
from sea level in southeastern Brasil to 1,700 m in Bolivia. It has been 
collected during every month but May, but in the southern part of its 
range is found only during the summer months. 

Diagnosis. The male may be distinguished from most other South Amer- 
ican species of Agapostemon by the acute lateral angle on its pronotum and 
lack of a "button" on its last visible sternum (6th). It may be distinguished 
from A. chapadensis by the shorter distal stylus of its genitalia (Figs. 173- 
174) and by the lesser amount of black on the posterior of its hind femur 
(Figs. 171-172). 

The female cannot be distinguished from A. chapadensis (cf. Variation 
below) but it may be distinguished from other South American Aga- 
postemon by the acute upturned flange on its anterior mesoscutal margin 
and lack of basal yellow bands on its metasomal terga. 

Variation. When A. semimelleus and A. chapadensis were first de- 
scribed, the females were separated chiefly on the basis of metasomal color. 
Supposedly the female metasoma of A. semimelleus is honey-colored while 
that of A. chapadensis is black. However, I feel that this character is un- 
reliable because I have seen specimens where the anterior of the metasoma 
was honey-colored while the posterior was black. I have also seen specimens 
with a dark brown metasoma. 

Independent evidence leads me to believe that intra-specinc variation 
masks interspecific differences between the females of these two species. The 
males of A. semimelleus are far more abundant and widespread than those 
of A. chapadensis. Unless one assumes that there are great disparities be- 
tween the distributions and relative abundance of females and males, one 
must conclude that the metasoma of A. semimelleus females is frequently 
black. It is also possible that some females of A. chapadensis have brown or 
amber metasomas. At present there seems no solution to this problem. A 
similar problem exists with respect to the sculpturing of females of A. 
texanus and A. angelicus. 



Revision of the Bee Genus Agapostemon 527 

Description 
male (Figs. 123-124, 171, 174, 220) 

General color of head and mesosoma bright metallic green, metasoma 
banded with black and yellow. Head (Figs. 123-124): with pale amber 
pubescence on interocular area and vertex, becoming creamy on gena. 
(1) habrum opaque yellow with apical and lateral margins transparent 
amber. Two-thirds as long as wide, apex narrowly rounded, basal ridge 
broad. Small, shallow, scattered punctures on dorsal ridge. Fringed with 
stiff submarginal bristles apically. (2) Clypeus with large (sometimes very 
large) scattered punctures on yellow region, becoming contiguously punctate 
to rugose on upper margin. (3) Interocular area with small to moderately 
small, deep, contiguous punctures below ocelli, becoming shallower and 
medium-sized at level of antennal sockets and still shallower below antennae; 
supraclypeal protuberance with large, contiguous, moderately deep punc- 
tures. Pubescence between antennal sockets Yi- 7 /} as long as scape, becoming 
shorter laterally, dorsally and ventrally and sub-tomentose ventro-laterally. 
(4) Vertex with deep contiguous punctures anteriorly and laterally, becom- 
ing shallower and slightly larger between ocelli and transversely carinate 
posteriorly. (5) Gena with parallel, subcontiguous, short carinae extending 
dorso-laterally from just behind smooth shiny area on antero-ventral margin. 
Pubescence postero-medially about as long as scape, becoming shorter an- 
teriorly, laterally and dorsally. (6) Malar area yellow, very short and with 
minute appressed hairs. (7) Mandible yellow with apical x /i ferruginous. 
(8) Antenna brown to brown-black above, scape yellow below and some- 
times with yellow extending dorsally at base, pedicel and flagellum amber to 
yellow below. Mesosoma: pubescence creamy to yellowish, becoming 
golden to fulvus on mesonotum and metanotum. (9) Pronotum with lateral 
angle and posterior lobe acutely pointed, lateral angle projecting antero- 
laterally; with conspicuous sharp carina extending postero-ventrally from 
apex of lateral angle; with weak irregular rugae postero-laterally. Pubescence 
above lateral angle and posterior lobe golden to fulvus, yellowish to creamy 
below. (10) Mesoscutum with large conspicuous flange laterally on anterior 
margin extending antero-dorsally; punctures small, very deep and con- 
tiguous, becoming rugose antero-laterally. (11) Mesoscutellum with lateral 
and medial punctures like that of mesoscutum, becoming larger and less 
dense laterally; area on each side of mid-line shiny with scattered punctures. 
(12) Metanotum coarsely punctate to rugose medially, becoming moderately 
coarsely rugose laterally. (13) Mesepisternum coarsely rugose anteriorly, be- 
coming moderately coarsely rugose posteriorly. (14) Metepisternum moder- 
ately finely rugose with horizontal rugae longer and slightly more prominent 
than other rugae. (15) Propodeum with propodeal carina strong and con- 
spicuous; propodeal shield very coarsely and deeply punctate to very 



528 The University Science Bulletin 

coarsely rugose; dorsal area very deeply rugose anteriorly and medially, be- 
coming moderately rugose to coarsely punctate laterally; lateral area with 
small, deep, subcontiguous punctures anteriorly and centrally, becoming 
coarsely rugose adjacent to propodeal carina. (16) Wing hyaline or nearly 
hyaline, veins dark brown. (17) Tegula pale to dark transparent amber with 
broad submarginal yellow band anteriorly (partially obscured by pubescence) 
and yellow to dark brown on proximal margin. (18) Fore leg entirely 
yellow, or yellow with brown postero-proximally on tibia and postero-dorsally 
on distal y 3 of femur. (19) Middle leg yellow with brown postero-dorsally 
on tibia, usually postero-dorsally at apex of femur, and occasionally on coxa. 
(20) Hind leg (Fig. 171) yellow with upper l / 2 of coxa metallic, with 
trochanter brown basally and sometimes amber apically, with brown apically 
on femur and antero-dorsally on tibia. Metasoma: (21) Terga 2-6 
amber to brown-black on posterior l / 2 and yellow on anterior l / 2 (often 
partially concealed by overlap) ; tergum 1 varying from pale amber with 
brown band on posterior margin to brown-black with lateral yellow spot 
antero-dorsally, but most commonly intermediate in condition with antero- 
dorsal yellow band separating dark brown anterior and posterior regions. 
Very conspicuously punctate with subcontiguous punctures largest on 
tergum 1, becoming progressively smaller on posterior terga. Pubescence 
golden to fulvus anteriorly on tergum 1, laterally on terga 1-7 and dorsally 
on terga 5-7; very short dorsally on terga 1-4, golden to fulvus on yellow 
bands and dark brown to brown-black on dark bands. (22) Sterna 2-6 
yellow but with amber to brown band at posterior margin of sterna 2-5, 
narrow on sternum 2, becoming progressively wider on successive sterna 
3-5; sternum 1 amber with metallic tints basally. Sternum 6 with low 
proximo-medial ridge. Pubescence limited to scattered, moderately long, 
fulvus to golden hairs on exposed areas. (23) Genitalia (Figs. 174, 220) 
with very short ventral lobes bearing apical brush of large stiff hairs. Distal 
stylus little longer than width of medial plate. 

female (Figs. 129-130) 
General coloration of head and mesosoma bright metallic green, meta- 
soma honey-colored to black. Head (Figs. 129-130): bright metallic green; 
pubescence white below becoming pale yellow above; sculpturing as in A. 
atrocaeruleus. Mesosoma: bright metallic green, but with tegula, legs 
and wing venation amber to brown; pubescence yellow or amber above, 
becoming creamy below; sculpturing as in A. atrocaeruleus but dorsal and 
lateral propodeal rugae more irregular. Metasoma: honey-colored to 
black; pubescence creamy, very short and appressed dorsally, becoming 
moderately long and erect anteriorly, laterally and posteriorly, and with 
narrow band of white tomentum basally on terga 2-4; sculpturing as in 
A . atrocaeruleus. 



Revision of the Bee Genus Agapostemon 529 

Agapostemon splendens (Lepeletier) 

Halictus splendens Lepeletier 1841. Type $, Museum National D'Histoire Naturelle, Paris. 
Agapostemon aeruginosas Smith 1853. Type 2, British Museum (Natural History). 
Agapostemon splendens. Robertson, 1897. 

I have seen the type of Halictus splendens, which was loaned to me by 
Dr. S. Kelner-Pillault. The type of A. aeruginosas was examined by Cock- 
erell, and more recently by Michener, and both believe it is conspecific with 
A. splendens. 

Distribution (Fig. 19). Agapostemon splendens is the most common 
species of Agapostemon in the southeastern United States and along the 
Gulf Coast. It is also common in the sandy areas of the Central States and 
Michigan. It is found as far south as Veracruz, Mexico, as far north as 
southern Saskatchewan, Canada, and as far west as south-central Utah. In 
Canada females have been collected in May, June, July and August; males in 
August and September. In Kansas females have been collected from April 
through October and males from July through October. In Florida both 
sexes have been collected every month of the year. While more common at 
elevations of less than 2,000 ft. (610 m), A. splendens has been collected as 
high as 6,600 ft. (2,012 m) in the Great Smoky Mountains National Park 
in Tennessee and 3,700 ft. (1,127 m) near Wray, Colorado. 

Diagnosis. The male may be distinguished from other species by its 
toothed hind femora, the grooved basal ridge on its basitarsus, the apical 
stylus on its gonostylus, and its brown wings. The female may be dis- 
tinguished from other species by its metallic green metasomal terga, rounded 
lateral pronotal angle, punctate mesoscutum and brown wings. 

Description 

male (Figs. 71-72, 166, 179, 200) 

General coloration of head and mesosoma bright metallic green, meta- 
soma with black and yellow bands. Head (Figs. 71-72): pubescence white, 
yellowish on vertex. (1) Labrum as in A. texanus. (2) Clypeus with large, 
shallow, scattered punctures on lower l / 2 , becoming contiguous above. (3) 
Interocular area with deep, contiguous, medium punctures below ocelli, 
becoming rugose laterally and below; sculpturing of supraclypeal protuber- 
ance like that of adjacent clypeal area. (4) Vertex with deep, contiguous, 
medium punctures anteriorly, becoming rugose laterally, posteriorly and 
between ocelli. (5) Genu with short coarse carinae extending dorso-laterally 
from antero-ventral margin. (6) Malar area bright yellow, short. (7) Man- 
dible bright yellow, pale ferruginous on apical V4-V3- (8) Antenna black 
above with yellow sometimes present on basal x A- x /i of scape and present on 
apical flagellomere; scape yellow below, pedicel with yellow spot on under- 
side, and underside of flagellum pale amber. Mesosoma: pubescence 



530 The University Science Bulletin 

yellow to amber on mesonotum and metanotum, white elsewhere. (9) 
Pronotum with lateral angle rounded and rugose above; postero-lateral area 
with large vertical carina just anterior to several irregular horizontal carinae. 
(10) Mesoscutum with coarse contiguous punctures becoming rugose antero- 
laterally. (11) Mesoscutellum with coarse contiguous punctures anteriorly, 
becoming rugose posteriorly. (12) Metanotum rugose. (13) Mesepisternum 
coarsely rugose anteriorly, only slightly less coarse posteriorly. (14) Metepi- 
sternum with very coarse irregular horizontal carinae or coarsely rugose. 
(15) Propodeum with propodeal carinae distinct; propodeal shield coarsely 
rugose; very coarsely rugose postero-laterally, becoming rugulose antero- 
laterally; coarsely rugose dorsally with triangular medial area depressed, 
rugulose. (16) Wing yellowish brown, conspicuously darker on distal x A- x /i\ 
veins and pterostigma amber to brown with radius conspicuously darker 
than other veins. (17) Tegula pale transparent amber with green tints 
antero-basally and short transverse yellow band on anterior area curving 
rearward and almost reaching distal margin. (18) Fore leg yellow with 
coxa metallic green and trochanter with amber tints dorsally. (19) Middle 
leg yellow with metallic tints on coxa and dark brown streak centrally on 
postero-dorsal surface of tibia. (20) Hind leg (Fig. 166) yellow; coxa and 
trochanter metallic green; femur with brown on apical x A- x /i of anterior, 
dorsal and posterior surfaces; tibia with brown posteriorly on distal l / 2 , 
antero-dorsally on proximal l / 2 , basally, and apically on anterior and dorsal 
surfaces; basitarsus with amber on grooved portion of basal ridge and on 
apical groove. Metasoma: (21) Terga brown to brown-black with yellow 
bands on basal halves of terga 2-6, and with central yellow band on 
tergum 1; anterior face of tergum 1 pale amber to brown. Pubescence short, 
minute and dense on dorsal areas, brown-black on brown surfaces and white 
on yellow or amber surfaces; pubescence longer on anterior surface of 
tergum 1 and ventro-laterally on terga 5-7. (22) Sterna brown apically with 
varying amounts of yellow basally (always some yellow present); low, 
transverse, subapical ridge on sternum 4 (and sometimes 3) largest laterally; 
long, white, scattered pubescence on exposed areas. (23) Genitalia (Figs. 
179, 200) with ventral lobes large and fringed with hairs distally. 

female (Figs. 69-70) 

General coloration of head, mesosoma and metasoma bright metallic 
green to blue-green. Head (Figs. 69-70): pubescence white to yellowish 
on clypeus and intcrocular area, becoming yellow to amber on vertex; white 
on gena. (1) Labium as in A. texanus. (2) Clypeus with large scattered 
punctures below, becoming subcontiguous on upper %. (3) Interocular area 
rugose with coarse contiguous punctures below ocelli; supraclypeal pro- 
tuberance with coarse punctures and faint horizontal carinae. (4) Vertex 



Revision of the Bee Genus Agapostemon 531 

with coarse contiguous punctures anteriorly, becoming rugulose laterally 
and between ocelli to rugose posteriorly. (5) Gena with coarse carinae ex- 
tending postero-dorsally from antero-ventral margin. (6) Malar area very 
short, dark brown. (7) Mandible yellowish to amber with apical YrYz 
ferruginous. (8) Antenna dark brown to brown-black with underside of 
flagellum slightly paler. Mesosoma: pubescence white, becoming yel- 
lowish on mesonotum and metanotum. (9) Pronotum with lateral angle 
rounded; dorsal aspect of lateral angle and posterior lobe shallowly rugose; 
postero-lateral area with several carinae extending postero-ventrally and 
antero-dorsally. (10) Mesoscutum with coarse, deep, contiguous punctures 
becoming rugose antero-laterally (punctures sometimes appearing to be of 
two sizes as in A. texanns but surface never as shiny). (11) Mesoscutdhtm 
with large scattered punctures and with much smaller contiguous punctures; 
central region just lateral to mid-line often shiny and devoid, or nearly 
devoid, of small punctures. (12) Metanotum rugose to rugulose. (13) 
Mesepisternum coarsely rugose. (14) Metepistemum rugose or with very 
irregular horizontal carinae. (15) Propodeum with strong propodeal carina; 
propodeal shield coarsely rugose or with irregular carinae extending dorso- 
lateral^ from medial groove. (16) Wing as in $ . (17) Tegula as in 6 
but slightly darker amber. (18) Fore leg dark brown to brown-black; coxa 
metallic green; femur with pale yellow anteriorly, dorsally and posteriorly 
at apex; tibia with basal yellow spot dorsally and pale brown on anterior 
surface; tarsus pale brown; pubescence white to amber. (19) Middle leg 
like fore leg but with coxa largely brown and with pubescence brown-black 
dorsally on tibia and basitarsus. (20) Hind leg brown to brown-black, coxa 
with metallic tints dorsally; pubescence white to amber, becoming brown- 
black dorsally on tibia and basitarsus. Metasoma: (21) Terga bright me- 
tallic green to blue dorsally, becoming brown ventro-laterally and brown 
basally on tergum 1. Pubescence white, becoming black on terga 5-6; long 
ventro-laterally on anterior surface of tergum 1 and dorsally on terga 5-6; 
terga 2-5 with basal bands of white tomentum. (22) Sterna brown to brown- 
black with long white hairs scattered on exposed areas. 

Agapostemon swainsonae Cockerell 

Agapostemon swainsonae Cockerell 1910b. Type $, British Museum (Natural History). 

The type has been examined by C. D. Michener. Its shiny mesoscutum 
leaves no doubt as to its identity. 

The female is recognized here for the first time. 

Distribution. This species is apparently limited to the island of Jamaica. 
I have seen specimens from the following localities: Yardley Chase, Santa 
Cruz Mtns., St. Elizabeth, 1,500 ft., (2 S , Oct.) ; Mandeville (15); Port- 
land (2 9); Christina, 3,000 ft. (1 5 , July); Hanover, Great River District 
(19, Jan.) ; Baron Hill, Trelawny (19, June) . 



532 The University Science Bulletin 

Diagnosis. The male is easily distinguished from all other West Indian 
species by its very shiny, weakly sculptured mesoscutum. The female has 
not been recognized previously owing to its similarity to A. viridulus. It may 
be distinguished from all but A. viridulus and A. hispaniolicus by its bright 
metallic green head and mesosoma and black metasoma; from A. hispanioli- 
cus by its more rugose interocular area; and from A. viridulus by the lack of 
metallic green tints at the base of its mandibles and by the lighter color of 
its mandibles. 

Description 
male (Figs. 113-114, 147,209) 

General coloration of head .and mesosoma bright metallic green and 
shiny, metasoma amber to yellowish brown. Head (Figs. 113-114): pu- 
bescence yellowish on clypeus, interocular area and vertex; white on gena. 
(1) Labrum as in A. viridulus but yellow to white on basal ]/ 2 and pale 
transparent amber on apical l / 2 . (2) Clypeus yellow to white below and 
nearly as elongate as in A. viridulus; punctures scattered and with numerous 
transverse rugulae on metallic area; pubescence short and sparse. (3) 
Interocular area shallowly rugulose around antenna! sockets, becoming finely 
and shallowly punctate dorsally and ventro-laterally; supraclypeal protuber- 
ance with punctures and transverse rugulae similar to those on clypeus; 
pubescence as in A. viridulus. (4) Vertex with very shallow, small, scattered 
punctures anteriorly and laterally, becoming finely rugose posteriorly; area 
between ocelli almost impunctate; pubescence as in A. viridulus. (5) Gena 
as in A. viridulus but with rugae weaker and more numerous. (6) Malar 
area as in A. viridulus but slightly shorter. (7) Mandible as in A. viridulus. 
(8) Antenna medium brown above, scape yellow to white below, flagellum 
amber to creamy below, underside of pedicel slightly darker than underside 
of flagellum. Mesosoma: pubescence creamy to yellowish on mesonotum 
and mctanotum, white laterally, ventrally, and on propodeum. (9) Pro- 
notum as in A. viridulus. (10) Mesoscutum very shiny (more so than in any 
other species of Agapostemon); very shallow, small punctures separated by 
their diameters. (11) Mesoscutellum very shiny with very small, shallow, 
widely scattered punctures. (12) Metanotum shiny, with minute, shallow, 
scattered punctures medially, becoming finely and irregularly rugulose later- 
ally. (13) Mesepisternum as in A. viridulus but sculpturing finer and shal- 
lower. (14-15) Metepisternum and propodeum as in A. viridulus but with 
slightly shallower sculpturing. (16) Wing as in A. viridulus (1 S with white 
pigment on underside of pterostigma and veins). (17) Tegula pale trans- 
parent, metallic tints absent or very faint, and yellow to creamy transverse 
lunule truncate basally and almost reaching distal margin. (18-19) Fore and 
middle legs yellowish amber with strong metallic tints on coxae; pubescence 
white on coxae, amber on other segments. (20) Hind leg (Fig. 147) as in 



Revision of the Bee Genus Agapostemon 533 

A. viridulus but with brown areas much paler brown (or amber). Meta- 
soma: (21) Terga amber with indistinct, slightly darker bands centrally. 
Pubescence moderately long, scattered, white ventro-laterally and on anterior 
surface of tergum 1; very short, inconspicuous white tomentum basally on 
terga 2-6 (usually hidden by overlapping terga) ; inconspicuous, dense, short, 
prostrate hairs dorsally on terga 1-4; long fuliginous hairs dorsally on terga 
4-7. (22) Sterna amber to yellowish with long, white, scattered hairs on ex- 
posed areas. (23) Genitalia as in A. viridulus but with only about l / 2 as 
many lateral grooves on gonocoxa (Fig. 209). 

female (Figs. 35-36) 

As described for A. viridulus but slightly bluer on head and metasoma; 
with finer sculpturing; with lower portion of clypeus, basal % of mandible, 
and labrum amber; and without metallic tints basally on mandible. 

Agapostemon texanus Cresson 

Agapostemon texanus Cresson 1872. Type 9, Academy of Natural Sciences, Philadelphia. 

Agapostemon texanus subtilior Cockered 1898. Type 9 (?), whereabouts unknown; neotype 
9 , University of Colorado Museum, Boulder. 

Agapostemon borealis Crawford 1901. Type 9, Academy of Natural Sciences, Philadelphia. 

Agapostemon calijornicus Crawford 1901. Lectotype 9, U.S. National Museum. 

Halictus (Agapostemon) brachycerus Vachal 1903 (new synonymy). Type $, Museum Na- 
tional D'Histoire Naturelle, Paris. 

Agapostemon texanus iowensis Cockerell 1910c (new synonymy). Type 9, U.S. National 
Museum. 

Agapostemon proscriptus Cockerell 1912b (new synonymy). Type 9> U.S. National Museum. 

Agapostemon joseanus Friese 1916 (new synonymy). Type 9, Zoologisches Museum, Hum- 
boldt Universitat, Berlin. 

Agapostemon suljuripes Friese 1916 (new synonymy). Type $, whereabouts unknown. 

Agapostemon cyanozonus Cockerell 1924 (new synonymy). Type 6 , California Academy of 
Sciences, San Francisco. 

Agapostemon proscriptellus Cockerell 1924 (new synonymy). Type 9, California Academy of 
Sciences, San Francisco. 

Agapostemon texanus vandy\ei Cockerell 1925. Type 9, California Academy of Sciences, 
San Francisco. 

Agapostemon calijornicus psammobius Cockerell 1937b (new synonymy). Type $ , California 
Academy of Sciences, San Francisco. 

Agapostemon angelicas idahoensis Michener 1937 (new synonymy). Type 9, Snow En- 
tomological Museum, University of Kansas, Lawrence. 

Agapostemon calijornicus clementinus Cockerell 1939 (new synonymy). Type 9, University 
of Colorado Museum, Boulder. 

I have examined the types of all the nominal species and subspecies 
listed above except A. texanus subtilior and A. suljuripes. As I find the 
females of A. texanus are indistinguishable from those of A. angehcus, female 
synonymies are based upon the distributions of the males (see map, Fig. 20). 
The type localities are as follows: A. texanus, Texas; A. borealis, Vancouver 
I., Canada; A. calijornicus, Pacific Grove, California; H. (A.) brachycerus, 
Guatemala; A. texanus iowensis, Ames, Iowa; A. proscriptus, Guatemala 
City, Guatemala; A. joseanus, San Jose, Costa Rica; A. cyanozonus, 



534 The University Science Bulletin 

Guaymas, Mexico; A. proscriptellits, Guaymas, Mexico; A. texanus vandy\ei, 
Yosemite Valley, California; A. californicus psammobius, San Miguel I., 
California; A. angelicas idahoensis, Challis, Idaho; A. californicus demen- 
ti mis, San Clemente I., California. 

The type of Agapostemon texanus subtilior Cockerell has not been 
identified by subsequent revisors and it is possible that Cockerell failed to 
designate the type series with appropriate labels. I have designated a female 
in the University of Colorado Museum as the neotype. This specimen is 
probably from the series described by Cockerell as it bears a label identical 
to that of some of the females mentioned by Cockerell, "Pasco, Wash/5 
25 96." 

I have not been able to examine a type of Agapostemon sulfuripes 
Friese, but the description is good and its identity with A. texanus is obvious. 

Distribution (Fig. 20). The range is greater than that of any other 
species in the genus. It may be found from southern Canada (52° N) to 
central Costa Rica (10° N) and from Cape Cod, Massachusetts (70° W) 
to Vancouver Island, Canada (125° W). Agapostemon texanus is most 
abundant on the western coast of the United States; fairly abundant in the 
United States west of the Mississippi River, except for the arid Southwest; 
uncommon in the arid Southwest, the United States east of the Mississippi 
River, Mexico, and Central America. This species occurs from sea level on 
the Adantic, Gulf and Pacific coasts to about 8,000 ft. (2,438 m) in the 
Sierra Nevada of California. At the northern limits of its range (British 
Columbia) females of A. texanus have been collected from May through 
September and males from July through October; in Kansas females (in- 
cluding those of A. angelicus) have been collected from April through 
November and males in April (very rare) and from June through October; 
in Arizona females (including those of A. angelicus) have been collected 
from February through November and males in July and August; and in 
Mexico females (including those of A. angelicus) have been collected from 
February through October and in December, and males from April and 
June through September. 

From its geographic distribution it appears that A. texanus is able to live 
under a wide variety of climatic conditions and in 70 of the 116 plant com- 
munities (see Appendix A) occurring in the United States (Kiichler, 1964). 
Considering the lack of apparent climatic, biotic or edaphic factors correlated 
with its distribution, I find the distribution of A. texanus inexplicable. In a 
widespread and heterogeneous as this (cf. discussion of variation), 
> which limit its distribution in one area may be relatively unimportant 
in another. 

is. The male may be distinguished from males of many other 
5 toothed hind femora, the apical stylus on its gonostylus, and 



Revision of the Bee Genus Agapostemon 535 

the lack of a low medial ridge on the apical l / 2 of its last visible sternum; 
from A. splendens by the lack of a grooved basal ridge on its hind basitarsus; 
and from A. angelicas by the apical stylus of its gonostylus (Figs. 180-181) 
and usually by the presence of a black stripe anteriorly on its hind tibia (if 
absent, then also lacking black stripe on posterior surface of hind tibia). 
The female may be distinguished from most other species by its bright 
metallic green to blue metasomal terga, its almost hyaline wings, and its 
two distinct sizes of mesoscutal punctures. I have not distinguished the 
females of A. texanus from those of A. angelicits except where they occur 
well outside the distribution of A. angelicas males (Fig. 20). 

Variation. The amount of variation is reflected by the length of the 
synonymy, which consists of 15 names. Although specimens from different 
parts of the range differ in color, pattern and sculpturing, the male genitalia 
are all alike and there are intergrades among the other characters. There- 
fore, I conclude that the described variants all belong to one species. A. 
angelicas is superficially more similar to the typical form of A. texanus than 
is A. texanus vandy\ei. However, I regard A. angelicits as a separate species 
because: the male genitalia — which provide excellent specific characters in 
many other species of Agapostemon — are different from those of A. texanus 
(Figs. 180-181); there are no intergrades between the two species with 
regard to certain color patterns; the two are sympatric. 

The variation within A. texanus falls into three major categories. These 
are the area and density of melanic pigmentation, the amount and hue of 
structural (metallic) coloration, and the sculpturing. 

Melanic pigmentation is best seen in males, as it is obscured or replaced 
by the structural colors predominant in females. The melanic pigmentation 
varies in density from pale amber to brown-black, and varies both within 
and among localities. It was expected that there might be concordance of 
variation in the size of melanic maculations on the appendages. However, 
specimens from San Clemente Island, California have no yellow (0%) on 
either the fore coxa or lower l / 2 of the scape; whereas specimens from Jalisco, 
Mexico, have little (10%) yellow on the fore coxa but much (95%) on the 
lower Y z of the scape. Nor could precipitation, temperature or other ecolog- 
ical factors account for the similarity between, for example, populations 
from New Hampshire and San Clemente Island, California (underside of 
scape 100% black) or between Saskatchewan, Canada, and Missouri (under- 
side of scape 100% yellow). 

There are considerable differences of color pattern among various popula- 
tions (Table 1). However, they can all be explained in terms of an increase 
or decrease in the amount (area) of melanic pigmentation. For example, 
although the black stripe on the posterior surface of the hind femur is 
absent in some populations and present in others, it can be shown that this 



536 



The University Science Bulletin 






03 



to 
to 

re 



re 
O 



c 
re 

<u 

a 
re 



c 

Ui 

a, 
1-1 






3 2 



8 






re 



^ 3 



. 


rt 


^ 


(-i 




6 


<-w 


c 


o 


i- 




M-l 


to 


CV5 


a 


1-1 


o 


o 


• ~* 




4-» 


- 


re 


U 






3 

a, 


"re 


o 


3 


Dh 


t~* 




u 


-a 

V 


3 
u 

4— » 


4— > 

u 


co 


JJ 




u 




to 




tJD 




a 




o 




B 




re 




c 




o 




■ 1—1 




4-» 




re 








1h 




re 




> 




u 




us 





3 
O 



^ co 

£ a 

£ E 

n 



21 

t/5 

re 





rt 


rt 


£ 


"O 


U 


rt 


-C 


C 


o 


c3 


rt 




(A 




rt 




CO 


a 




o 




So 


_y 



o 



►J 

ft 

o 



rt o 

•5 O 



.^ c 

rt "C 



n 




i — i 








'a 




u 


o 


c 


a 


'-t-H 


rt 


QJ 




00 


E 


"rt 




OJ 


u 




u 



C 


o 


u 


u 


X 


vs 


flj 




•^ 


rt 


^. 


1 



00 



o 



IA 

ON 

I 



o 

I 

CM 

I 

in 



CM 

I 
O 

a\ 






o 
I 

"-N 

I 
O 



-2«o 



•^ 



<u rt 



o 
■q- 

I 

"? 

o 



on 
I 



o 

7 

o 



o 

I 
cs 

"-N 

I 

o 



7 

o 

ON 



C5 



On 

I 
o 

o 



rt "« 

X rt 

O tu 

o ™ 

O cN 

to, w 



l/N 

I 

<=> 

°? 

o 

ON 



o 
l/N 

I 
<=> 

°P 

in 

on 



o 

o 



I 
"A 

7 

o 



o 

7 

>n 

7 

O 



o 

I 

°p 

in 

ON 



-5 
o *o ~ 

"< O i-, 
'S u ° 

■o 3 n 

•C Ji V4-, 

" "« „ 



o 

H-N 

IX 

I 

o 

o 



o 

I 

T 

o 
o 



1) 




c 


rt 




u 


"3 
u 


JO 


96 




s? 








C 




i> 



4-J 

.2 
-a 



IE 

00 



o 

I 

"N. 

I 

o 
in 



T 

o 



5^ 



3 

a 



3 

Q 



o 


o 


1 


1 



3 

Q 



J3 



o 



w 3 

.2 S 

U. M-4 

O .3 



CH- 

"rt o+ 

E 5 o 



O C 



3 
O 


jo aj 


u 


S^ 


rt 


4J C 


CO 

u 






3 
o 

u 



c 

o+ y 

(O 

E g, 

5 « 

3 O 
O) u, 

o .2 

to O 



s 

u 

CQ § 



o 



— 





V5 


a 


>> 


3 
O 


QJ 


c 




9 o 


IE 


rt 


CQ — 


00 


t/3 


CO ^ 




u 


-S 




H 


w 



CO 


C 


3 


u 


o 


u 


OJ 

tJ 


Jh o 


rt 

4-> 
CO 


O-B 



PQ CO 



o 



z 




_o 


?, "a 


"3 





o+ 
o+ 

co r^i 

£ Z 

o —i 

cj 

— >N 

rt rt 

C (50 

J3 (5 
00 ^-^ 



Revision of the Bee Genus Agapostemon 537 

stripe is absent only in those populations in which the anterior stripe is 
reduced or absent, and that the reverse is never true. Every specimen in 
which the posterior hind femoral stripe is present and the anterior stripe 
absent proves, upon examination of the genitalia, to be A. angelicas rather 
than A. texanas. Therefore, although differences of color pattern do exist 
among individuals and populations, they are a result of the amount of 
melanic pigmentation. 

One of the most obvious variables is the hue of the structural (metallic) 
coloration. It is usually pure green, but in some populations is yellowish or 
blue. The extremes of the yellowish and blue forms are so distinctive that 
they may be distinguished at a distance of several feet. The yellowish form 
(A. vandy\ei) occurs in the Sierra Nevada of California and in the Coast 
Ranges north of San Francisco. Limited to montane areas, it is not found 
along the coast or in the Central Valley. This form has strong brassy yellow 
overtones which are most striking in the Sierra Nevada south of Yosemite 
National Park and in the southern portion of the Coast Ranges north of 
San Francisco. The yellow cast becomes less pronounced further north to 
the point where it is replaced by the pure green of A. texanus texanus along 
the California-Oregon border. 

The blue form of A. texanus (A. californicus) is found on the coast of 
California south of Cape Mendocino, in the Central Valley, and on the 
islands off the coast of southern California. It occurs only at low elevations 
and is not sympatric with the yellow form. It is most easily recognized by 
the bluish (rather than green) cast of the metallic coloration, which is 
particularly pronounced in the mesoscutum of males. This hue is most 
intense in the southern portion of the range, most notably on San Clemente 
Island (A. calijornicus clementinus), where it is a very dark violaceous blue. 
However, it becomes greener further north to the point where, along the 
California-Oregon border, it is replaced by the pure green of A. texanus 
texanus. 

It is interesting that there do not seem to be intermediates between the 
yellow and blue forms in the southern portions of their ranges. It would be 
logical to regard them as separate taxa were it not for the green intermediate 
forms to the north. Although the origins of this geographic distribution of 
color forms are doubtless environmental, the populations must be relatively 
homogeneous genetically for this character because annual or local environ- 
mental fluctuations have no apparent effect upon the colors of the popula- 
tions. 

Variation in sculpturing is best seen on the mesosoma of females. The 
punctures on the mesoscutum range from deep and contiguous (Fig. 24) to 
relatively shallow and far apart (Fig. 23). The former condition produces 
a rough and dull appearance, the latter a relatively smooth and shiny surface, 



538 The University Science Bulletin 

There are slight geographical variations in this character (Table 1), but 
the variation within a given locality is considerable. Although the rugae on 
the propodeum all follow the same general pattern, they vary considerably 
in number, coarseness, angle of divergence and number of interconnections 
(Figs. 25-26). This variability is so great within a locality that it is im- 
possible to discern a significant difference among localities. In fact it is 
so great as to invalidate the contention of Sandhouse (1936) and her pre- 
decessors that A. texanus females may be distinguished from those of A. 
angelicits by propodeal sculpturing. 

Description 
male (Figs. 75-76, 165, 180, 201) 

General coloration of head and mesosoma bright metallic green to blue, 
metasoma with black and yellow bands. Head (Figs. 75-76): (1) Labritm 
opaque lemon yellow with transparent pale amber margins (distal % amber 
on specimens from San Clemente I., California). Nearly as wide as long; 
distal margin rounded; and large transverse ridge on basal l / 3 with very 
slight medial depression. Punctures on crown of basal ridge shallow and 
separated by 2-3 times their diameters; submarginal punctures large and 
shallow medially, becoming smaller laterally and disappearing just anterior 
to basal ridge. Pubescence on crown of basal ridge minute, simple and erect; 
simple hairs in submarginal punctures about three times as long as puncture 
is wide and deflected distally. (2) Clypeus with coarse punctures contiguous 
at upper margin, becoming subcontiguous at lower margin and becoming 
smaller and shallower laterally; pubescence white, dense and about 2/5 
length of clypeus medially, decreasing to 1/5 length of clypeus and loosely 
appressed at lateral margin. (3) Interocular area with small, deep, contiguous 
punctures below ocelli, becoming coarser laterally and moderately deeply 
rugulose above antennal sockets and shallowly rugulose ventro-laterally; 
supraclypeal protuberance coarsely punctate to shallowly rugose; pubescence 
white, erect, l / 2 as long as scape medially, becoming l / 4 as long laterally. 
(4) Vertex with small, deep, contiguous punctures anteriorly, becoming 
shallower laterally; weakly rugose posteriorly, becoming weakly rugulose to 
weakly punctate between ocelli; pubescence white, % as long as distance 
between lateral ocelli medially, decreasing to l / 2 as long as distance between 
lateral ocelli laterally. (5) Gena with hypostomal carina about 2.5 times as 
long as width of proboscidial fossa; short parallel carinulae extending 
postero-dorsally from antero-ventral margin; pubescence white, ranging in 
length from /5 as long as hypostomal carina centrally to 1/5 as long 
laterally and dorsally. (6) Malar area yellow to amber; very short; pubes- 
cence short, white, appressed. (7) Mandible opaque lemon yellow with 
distal y 3 ferruginous; punctures few, shallow and scattered thinly over basal 



Revision of the Bee Genus Agapostemon 539 

Y 2 ; pubescence scant with small patch of short white hairs on base of man- 
dible. Row of approximately eight evenly spaced amber hairs along lower 
margin of basal 2 / 3 of mandible, with proximal hairs nearly twice as long 
as distal hairs. (8) Antenna brown to brown-black above with pale amber 
or yellow on apical l / 2 of apical flagellomere; flagellum pale amber below; 
scape and pedicel yellow to black below (see remarks on Variation). 
Mesosoma: pubescence white, becoming creamy to golden on mesono- 
tum and metanotum; metallic coloration brassy green to blue. (9) Pronotum 
with lateral angle and posterior lobe slightly angular. Weak carina extend- 
ing postero-ventrally from lateral angle; low carina extending postero- 
ventrally from center of lateral portion; weak horizontal carinulae near 
postero-lateral margin; long pubescence above lateral angle and posterior 
lobe and very fine white tomentum on postero-lateral area. (10) Meso- 
scutum with numerous small, dense, deep, contiguous punctures becoming 
rugose anterolaterally. (11) MesoscuteUum with punctation like that of 
mesoscutum but shallower, less dense, and often with shiny lateral area; 
pubescence like that of mesoscutum. (12) Metanotum coarsely punctate to 
rugulose; pubescence like that of mesoscutum. (13) Mesepistermtm coarsely 
rugose anteriorly and only slightly less coarsely rugose posteriorly; pubes- 
cence long. (14) Metepisterniim coarsely rugose, sometimes with horizontal 
rugae more prominent than others; pubescence long. (15) Propodeum with 
propodeal carina moderately weak to weak; propodeal shield moderately 
coarsely to moderately finely rugose; dorsal area moderately coarsely rugose 
medially and anteriorly, becoming finely rugose to finely rugulose postero- 
laterally; lateral area moderately finely rugose posteriorly, becoming punctate 
anteriorly. Pubescence very short dorsally, moderately long laterally and 
posteriorly. (16) Wing nearly hyaline, with pterostigma and all veins but 
radius amber; radius dark brown. (17) Tegula dark to very pale trans- 
parent amber, with large pale yellow lunulate maculation parallel to — but 
separated from — anterior margin, with metallic tints on basal margin, and 
with short pale yellow band on posterior margin. (18) Fore leg with coxa 
metallic green; trochanter yellow to brown-black; femur entirely yellow 
to yellow with dark brown on all but distal portion of posterior surface; 
tibia entirely yellow to yellow with dark brown on all but distal l / 3 of 
ventral surface and short brown stripe sub-basally on dorsal surface; tarsus 
yellow. Pubescence white, becoming pale amber on tarsus. (19) Middle leg 
with coxa brown-black tinted with metallic green; trochanter yellow with 
brown spot basally on posterior surface to entirely brown-black; femur en- 
tirely yellow to yellow with brown-black stripe broadest basally and narrowest 
distally extending from base to apex on posterior surface; tibia yellow with 
dark brown sub-basal stripe on basal l / z of dorsal surface to yellow with 
brown-black sub-basal stripe on basal % of dorsal surface and brown-black 



540 The University Science Bulletin 

stripe entire length of ventral surface; tarsus yellow. Pubescence similar to 
that on fore leg. (20) Hind leg (Fig. 165) with coxa bright metallic green; 
trochanter brown-black with metallic tints to yellow with dark brown 
basally; femur yellow with large brown apical spot postero-dorsally and 
sometimes with basal brown spot posteriorly; tibia yellow with small brown- 
black area at base and usually with apical streak of brown to brown-black 
on posterior surface and sub-basal streak of brown to brown-black antero- 
dorsally (posterior streak or both posterior and antero-dorsal streak may be 
absent) ; tarsus yellow. Basal ridge on basitarsus low and inconspicuous. 
Pubescence white, usually pale amber posteriorly on tarsus. Metasoma: 
(21) Terga black with yellow bands on basal halves of terga 2-5 and centrally 
on tergum 1; ventro-lateral margins of terga 1-6 transparent amber; metallic 
tints apically on terga 4-5, pygidium amber to yellow. Pubescence dorsally 
on terga 1-4 very short and inconspicuous, pale on yellow band, dark on 
brown-black bands; moderately long and white anteriorly on tergum 1, 
laterally on terga 3-5, and dorsally and laterally on terga 5-7. (22) Sterna 
yellow with brown band apically on sterna 2-5 (narrow on sternum 2 but 
occupying most of sternum 5) ; sternum 6 brown to yellow with brown band 
basally; sternum 1 yellow to amber, with metallic tints basally; sternum 4 
with low, transverse ridge reaching distal margin laterally and usually with 
faint metallic spot medially. Moderately long, scattered, white hairs on ex- 
posed areas and with 2-4 large stout bristles disto-laterally on sternum 4. 
(23) Genitalia (Figs. 180, 201) with apical stylus, medial plate and basal 
stylus present; ventral lobe moderately large, fringed with short hairs 
distally. 

female (Figs. 21-26, 67-68) 

General coloration bright metallic green to blue or brassy. Head (Figs. 
21-22, 67-68). (1) Labrum dark amber to ferruginous. Basal ridge prom- 
inent, proximal face sloping gently toward proximal margin, distal face 
sharply declivous, and apex of ridge with deep punctures; distal portion 
narrow, abruptly rounded at apex; conspicuous median keel on antero-distal 
portion, continued basally almost to basal ridge; anterior margin of keel 
flattened and slightly flared laterally; single row of dark amber, wide, flat, 
stiff bristles curving anteriorly at their apices and forming fimbria on antero- 
lateral margins of labrum. (2) Clypeus brown-black below, becoming green 
above; large, scattered, deep punctures below, becoming slightly smaller and 
subcontiguous above; pubescence creamy to amber with single row of amber 
bristles curving downwards from just above lower margin of clypeus. (3) 
Interocular area with deep, moderate sized, contiguous punctures below 
ocelli becoming rugose laterally and below; supraclypeal protuberance with 
moderate sized, scattered punctures between short transverse rugae. (4) 



Revision of the Bee Genus Agapostemon 541 

Vertex with deep, small to moderate sized, contiguous punctures anteriorly, 
becoming smaller and shallower laterally; rugulose posteriorly and between 
ocelli; pubescence creamy to amber. (5) Genu (Fig. 22) with regular, 
moderately coarse carinae extending postero-dorsally from antero-ventral 
margin; pubescence white and long. (6) Malar area amber to brown-black; 
very short; pubescence very short, sparse, white and tomentose. (7) Mandible 
yellow to pale amber with apical V^-Yz ferruginous; moderately long, white, 
scattered pubescence at base and single row of moderately long amber pu- 
bescence on lower margin. (S) Antenna dark brown to brown-black with 
underside of flagellum amber. Mesosoma: pubescence white, becoming 
creamy to amber on mesonotum and metanotum. (9) Pronotum as in $ 
but with lateral angle and posterior lobe slightly more angular. (10) 
Mesoscutum (Figs. 23-24) with medium sized, moderately deep punctures 
separated by 2-3 times their diameters and interspersed with more numerous 
smaller and shallower contiguous to subcontiguous punctures; rugose antero- 
laterally (see discussion of Variation); pubescence moderately long. (11) 
Mesoscutellitm with punctation like that of mesoscutum but small punctures 
smaller, more numerous and shallower; often shinier than mesoscutum; 
pubescence moderately long. (12) Metanotum finely and shallowly rugulose; 
pubescence moderately long. (13) Me sepi sternum with coarse rugae an- 
teriorly, becoming slightly less coarse posteriorly; pubescence moderately 
long. (14) Metepistemum with coarse, interconnected, horizontal carinae; 
pubescence moderately long. (15) Propodeum (Figs. 25-26) with prominent 
propodeal carina; propodeal shield weakly rugose with stronger rugae ex- 
tending dorso-laterally from medial groove; dorsal area with weak, irregular, 
interconnected rugae extending posteriorly from anterior margin, becoming 
moderately finely rugose postero-laterally; lateral area with contiguous, 
horizontal carinulae and scattered small punctures; pubescence short dorsally, 
moderately long laterally and posteriorly. (16) Wing as in $ . (17) Tegula 
as in i but slightly darker amber. (18) Fore leg with coxa dark brown to 
brown-black and tinted with metallic green; trochanter dark brown to 
brown-black, often with narrow, inconspicuous, yellow band apically on 
posterior surface; femur brown to brown-black, with apical yellow band; 
tibia brown-black, with amber to yellow anteriorly; tarsus brown-black to 
pale amber; pubescence creamy on coxa, becoming amber on tarsus. (19) 
Middle leg similar in color to fore leg but with pubescence fuliginous dor- 
sally on distal portion of tibia and amber ventrally on femur and tibia. 
(20) Hind leg with coxa dark brown to brown-black with strong metallic 
green tints dorsally; trochanter brown to brown-black; femur, tibia and 
tarsus brown-black to pale amber; pubescence creamy on coxa, trochanter 
and femur, becoming amber on tibia and tarsus and dark fuliginous to 
black on dorsal surface of tibia (darkest basally). Metasoma: (21) Terga 



542 The University Science Bulletin 

metallic green to blue or brassy; white bands of pruinose pubescence cover- 
ing basal l / $ of terga 1-4 and white, very short, simple hairs covering dorsal 
% of terga 1-4; terga 5-6 with long, dark brown to black pubescence. (22) 
Sterna brown to brown-black with metallic tints basally on sternum 1; 
long, creamy, scattered hairs on exposed areas. 

Agapostemon tyleri Cockerell 

Agapostemon tyleri Cockerell 1917a. Type 9 , U.S. National Museum. 

Agapostemon martini Cockerell 1927, 9 only, (new synonymy). Type 9, California Academy 
of Sciences. 

I have examined the holotypes of Agapostemon tyleri and A. martini 
and find them conspecific. Contrary to the opinion of Sandhouse (1936) 
I do not find the type of A. tyleri conspecific with that of A. coloradensis 
Crawford (cf. Diagnosis). Cockerell correctly associated the sexes of A. tyleri 
but incorrectly associated the male of A. cocker elli with the female of A. 
martini. The male paratypes of A. martini are misidentified and belong to 
Agapostemon coc\erelli Crawford. 

Distribution (Fig. 27). Agapostemon tyleri occurs in the desert South- 
west and in arid parts of Mexico. I have seen specimens from as far north 
as Jerome, Arizona; as far south as Tehuacan, Puebla, Mexico; as far east 
as the Gulf Coast in Tamaulipas, Mexico; and as far west as Prescott, 
Arizona. It ranges from sea level in Tamaulipas, Mexico, to 8,200 ft. (2,499 
m) at Pachuca, Hidalgo, Mexico. Throughout most of its range it occurs 
at altitudes of over 4,000 ft. (1,219 m) on the central plateau. In the United 
States males have been collected from February through April, June through 
September, and November; and females in April and May, July through 
September, and November. In Mexico males have been collected in Feb- 
ruary, May through September, and November; and females from June 
through September, and December. 

Diagnosis. The male may be distinguished from all North American 
species except A. coloradinus by the lack of yellow maculations on its dark 
brown to black metasomal sterna; from A. coloradinus by its smaller size, 
by the black streak extending the entire length of the posterior surface of 
its hind tibia (Figs. 159-160), and by its broader penis valve, which lacks 
conspicuous hairs on its dorsal crest (Figs. 226-227). The female may be 
distinguished from many other species occurring north of Mexico by its 
black metasoma and by the lack of yellow on its clypeus; from A. virescens 
by its finer genal carinulae (2-3 per 0.25 mm in A. virescens and 5-6 per 0.25 
mm in A. tyleri, cf. Figs. 28-29); and from A. coloradinus by its colorless 
wings, darker tegulae, and shinier supraclypeal protuberance. This species 
is very close to A. coloradinus. 



Revision of the Bee Genus Agapostemon 543 

Description 
male (Figs. 95-96, 159, 184, 198, 227) 

General coloration of head and mesosoma bright metallic blue-green to 
blue, metasoma with black and pale yellow to creamy bands. Head 
(Figs. 95-96): pubescence snowy white. (1) Labrum as in A. texanus but 
more rounded at apex. (2) Clypeus as in A. virescens but punctures shal- 
lower and pubescence more dense. (3) Interocular area with coarse con- 
tiguous punctures (some specimens finely rugose around antennal sockets) ; 
supraclypeal protuberance as in A. virescens. (4) Vertex as in A. virescens. 
(5) Gena as in A. virescens but with carinae much finer. (6) Malar area 
yellow, never amber; short. (7) Mandible as in A. virescens. (8) Antenna 
as in A. coloradinus. Mesosoma: pubescence snowy white. (9) Prono- 
tum as in A. virescens but with weaker sculpturing. (10) Mesoscutum 
as in A. virescens but with finer punctures and with anterior margin punc- 
tate. (11-13) Mesoscutelhtm, metanotum and mete pi sternum as in A. 
virescens but with finer sculpturing. (14-17) Metepisternum, propodeum , 
wing and tegula as in A. coloradinus. (18-20) Fore, middle and hind legs 
as in A. virescens but posterior surfaces of tibiae and femora nearly com- 
pletely covered with dark brown to brown-black, and usually with small 
yellow rim apically on ventral surface of trochanters (Fig. 159). Meta- 
soma: (21-22) Terga and sterna as in A. coloradinus but with creamy 
band on tergum 1 almost always broadly interrupted medially. (23) Geni- 
talia (Figs. 184, 198, 227) as in A. virescens but with much smaller fold on 
medial plate of gonostylus and with prominent dorsal crest on penis valve. 

female (Figs. 29, 51-52) 

General coloration of head and mesosoma bright metallic blue-green, 
metasoma black with white hair bands. Head (Figs. 51-52): (1) Labrum 
as in A. texanus. (2-4) Clypeus, interocular area and vertex as in A. 
virescens but with center of supraclypeal protuberance shiny and nearly im- 
punctate. (5) Gena as in A. virescens but with much finer and more nu- 
merous (5-6 per 0.25 mm) carinulae. (6) Malar area very dark ferruginous 
to brown-black; very narrow, almost absent. (7-8) Mandible and antenna 
as in A. coloradinus. Mesosoma: snowy white pubescence sometimes 
very slightly fuliginous on mesonotum and metanotum. (9-15) Pronotum, 
mesoscutum, mesoscutellum , metanotum, mesepisternum, metepisternum 
and propodeum as in A. virescens but with sculpturing slightly shallower 
and finer. (16) Wing as in A. virescens but always clear, never slightly 
amber. (17) Tegula as in A. virescens but dark amber (also darker than in 
A. coloradinus). (18-20) Fore, middle and hind legs as in A. virescens but 
with pubescence slightly paler. Metasoma: (21) Terga as in A. virescens 



544 The University Science Bulletin 

but with pubescence on 5-6 paler brown. (22) Sterna as in A. virescens but 
darker. 

Agapostemon viequesensis Cockerell 

Agapostcmon tiequesensis Cockerell 1918b. Type 9 . U.S. National Museum. 

Agapostemon radiants portoricensis Cockerell 1919 (new synonymy). Type 8, U.S. National 

Museum. 
Agapostemon portoricensis. Wolcott, 1948. 

I have seen the types of both A. viequesensis and A. radiatus portoricensis 
and have no doubt that they are conspecific. 

Distribution. In addition to the types I have seen specimens from the 
following Puerto Rican localities: Aguada (1 8,9 Nov.), Arecibo (4 9, 
24-26 June; 1 8 & 7 9 , 30 July-1 Aug.), Caguas (4 8 , 28-29 May), Coamo 
Springs (2 6* , 28 Dec; 1 9 , 29 Dec), Dorado (1 8 , 15 July; 1 9 , 12 Apr.), 
Guayanilla (1 $ , 22 July), Manati (2 8 , 5 Mar.; 15,7 June), Mayaguez 
(1 8, Mar.; 2 9, 21-23 June; 1 9 , 16 Dec), Ponce (1 8, Jan.), Puerto 
Cangrejos (1 8 , 28 Jan.), Rio Piedras (1 8 , 23 Mar.), San Juan (3 8 , 11-14 
Feb.; 1 9 , 14 Apr.; 1 8 , 9-12 July), Vega Alta (1 9 , 27 Oct.) and Mona 
Island (1 8 & 1 9 , 21-26 Feb.; 2 8 , Aug.). I have also seen specimens from 
Nassau, New Providence I. (1 8 & 2 9 , 16 Dec.) and Arthurs Town, Cat 
I. (1 9,July-Aug.). 

Diagnosis. The male may be distinguished from A. centratus and A. 
kohhellus by its toothed hind femora; from A. viriditlus, A. obscuratus, A. 
hispaniohcus, A. swainsonae and A. cubensis by its lack of an elongate 
clypeal region; from A. cyaneus, A. sapphirinus and A. ochromops by its 
lack of milky to yellowish eyes and pterostigma; from A. insularis, A. 
jamaicensis and A. poeyi by its lack of metallic green or blue medially on 
metasomal tergum 3; and from A. columbi by its lack of metallic green 
tints medially on metasomal tergum 3. The female may be distinguished 
from other West Indian species with metallic metasomas by its lack of milky 
to yellowish eyes and pterostigma and by the tints at the base of its mandible. 

Description 
male (Figs. 117-118, 145, 187, 224) 

General coloration of head and mesosoma metallic blue-green, metasoma 
with brown-black and yellow bands. Head (Figs. 117-118): (1) Labrum 
yellow with apical margin transparent; basal ridge with very slight or no 
medial depression, apical l / 2 triangular; glabrous but for submarginal row 
"t simple bristles. (2) Clypeus green above, yellow below; punctures scat- 
tered and small on yellow area, becoming larger and more numerous on 
green area; short white scattered hairs on green area almost absent from 
yellow area. (3) Interocular area blue-green above, becoming yellow-green 
below; rugose, becoming coarsely, deeply and contiguously punctate above 



Revision of the Bee Genus Agapostemon 545 

antennal sockets; pubescence yellowish to creamy, % as long as scape be- 
tween antennal sockets, becoming shorter laterally and ventrally, and be- 
coming white and tomentose ventro-laterally. (4) Vertex with deep, con- 
tiguous, medium punctures anteriorly, becoming smaller and shallower 
laterally; rugose posteriorly, becoming irregularly rugulose between ocelli; 
pubescence as on interocular area but somewhat darker in color. (5) Gena 
with weak, fine, short rugulae extending postero-dorsally from antero-ventral 
margin; pubescence white, 1.5-2.0 times as long postero-medially as between 
antennal sockets but becoming shorter anteriorly, laterally and dorsally. 
(6) Malar area absent or very short. (7) Mandible yellow, apical / 3 ferru- 
ginous; single row of bristles postero-ventrally, longest basally, becoming 
l / 3 as long distally. (8) Antenna with scape yellow, becoming dark brown 
to brown-black on distal J4-/4 °^ u PP er surface; pedicel and flagellum dark 
brown to brown-black above, becoming pale amber to yellow below; apical 
flagellar segment yellow above and pedicel often slightly darker than flagel- 
lum. Mesosoma: pubescence wholly white or tinged with yellow on 
mesonotum and metanotum. ( 9) Pronotum with lateral angle and posterior 
lobe rounded; several to many fine, horizontal carinulae on postero-lateral 
area; pubescence on dorsal portions of lateral angle and posterior lobe as on 
mesoscutum, becoming very short, fine and inconspicuous on postero-lateral 
area. (10) Mesoscutum blue-green from above, yellow-green in oblique 
view; very finely and contiguously punctate, becoming rugose antero- 
laterally. (11) Mesoscutellum shinier than mesoscutum, with scattered 
large punctures interspersed with more numerous smaller and shallower 
punctures (faintly rugulose in some specimens). (12) Metanotum rugose. 
(13) Mesepisternum rugose anteriorly, becoming rugulose posteriorly and 
ventrally. (14) Metepisternum horizontally rugose. (15) Propodeum with 
distinct propodeal carina; coarsely rugose on propodeal shield, and posteriorly 
on dorsal portion, becoming moderately rugose antero-laterally; antero- 
dorsal margin with short rugae extending postero-laterally. (16) Wing 
transparent, slightly tinged with brown; radius brown-black, other veins 
and pterostigma brown. (17) Tegula pale transparent amber with metallic 
green tints antero-basally and yellow band almost reaching distal margin. 
(18-19) Fore and middle legs with coxae metallic green (only weakly tinted 
on middle leg), trochanters amber and remaining segments yellow with 
posterior of tibiae amber; pubescence pale amber. (20) Hind leg with coxa 
green, trochanter brown, femur yellow with brown dorsally on apical l / 3 , 
tibia dark amber to brown dorsally, becoming yellow ventrally, tarsus 
yellow; pubescence amber (Fig. 145). Metasoma: (21) Terga with 
yellow band on basal l / 2 of dorsal areas; anterior portion of tergum 1 amber; 
anterior l / 2 of dorsal areas brown, darkest anteriorly and becoming trans- 
parent on posterior margin; terga 4-5 often with very faint metallic green 



546 The University Science Bulletin 

tints laterally. Apex of pygidium rounded (Fig. 224). Pubescence very mi- 
nute, inconspicuous; white on yellow bands becoming brown on brown 
bands. Longer hairs white on anterior surface of tergum 1 and ventro-later- 
ally on terga 2-5, and becoming brown dorsally on terga 5-7. (22) Sterna yel- 
lowish to amber or pale brown with metallic green tints basally on sternum 
1. Sternum 4 with posterior margin slightly concave, and with low sub- 
marginal transverse ridge most prominent laterally; scattered white hairs on 
exposed areas. (23) Genitalia (Fig. 187) as in A. poeyi but apical stylus on 
gonostylus not as swollen at apex. 

female (Figs. 9, 11, 13, 15, 17, 39-40) 

General coloration of head,' mesosoma and metasoma bright metallic 
green to blue-green. Head (Figs. 15, 17, 39-40): pubescence white, be- 
coming fuliginous on vertex. (1) Labrum as in A. hispaniolicus. (2) 
Clypeus green above and brown to brown-black below; lower x / 2 with large 
punctures separated by at least their diameters, punctures on upper l / 2 
slightly obscured by faint transverse rugae; pubescence sparse, white, longest 
dorsally, shortest and least abundant ventrally, row of long amber bristles on 
lower margin. (3) Interocular area with large, deep, contiguous punctures 
centrally on upper % becoming moderately rugose laterally and on lower 
%; rugae nearest inner margins of compound eyes irregularly vertical in 
orientation. Pubescence Y 3 -y 2 as long between antennal sockets as scape, 
becoming shorter ventrally and laterally. (4) Vertex with large, deep, con- 
tiguous punctures anteriorly and laterally, becoming rugose posteriorly and 
rugulose between ocelli. (5) Genu with fine carinulae extending dorso- 
lateral^ from antero-ventral margin; pubescence postero-medially twice as 
long as pubescence between antennal sockets, becoming shorter anteriorly, 
laterally and dorsally. (6) Malar area brown, very short to absent. (7) 
Mandible amber with apical '/ 3 ferruginous; single row of hairs on postero- 
ventral margin long basally, becoming l / 3 as long distally. (8) Antenna 
brown-black above, slightly paler below; pedicel, proximal and distal 
flagellomeres often paler below than scape and remainder of flagellum. 
Mesosoma: (Figs. 9, 11, 13): pubescence white, becoming slightly fulig- 
inous on mesonotum and metanotum. (9) Pronotum with lateral angle 
and posterior lobe rounded; postero-lateral area with 3-6 weak, horizontal 
carinulae; pubescence above lateral angle and posterior lobe like that on meso- 
scutum; lateral surface with very fine, inconspicuous, short, white hairs. 
(10) Mesoscutum with very fine, contiguous punctures, becoming shallowly 
rugulose antero-laterally. (11) Mesoscutellum with scattered large, deep 
punctures interspersed with far more numerous very small, shallow punc- 
tures. (12) Metanotum rugose. (13) Me se pi sternum rugose with rugae 
finest posteriorly (specimens from Nassau, Bahamas with slightly coarser 



Revision of the Bee Genus Agapostemon 547 

sculpturing). (14) Metepisternum with regular horizontal carinulae. (15) 
Propodeum with strong propodeal carina; propodeal shield rugose or with 
irregular anastomosing rugae extending dorso-laterally from medial groove; 
dorsal area finely rugulose medially, laterally with anastomosing short carinae 
extending postero-laterally from anterior margin; lateral area with weak 
carinulae extending postero-dorsally from antero-lateral margin; pubescence 
very short and fine dorsally (absent medially), with long hairs posteriorly 
and laterally. (16) Wing as in £ . (17) Tegula dark amber to brown with 
green tints antero-basally with faint (often obscure) yellow band almost 
reaching distal margin. (18-20) Fore, middle and hind legs with coxae dark 
brown; with metallic tints on anterior surface of fore coxa and dorsal surface 
of hind coxa; trochanters brown, remaining segments amber; pubescence 
white to pale amber, becoming dark brown dorsally on middle tibia, hind 
tibia and apex of hind femur. Metasoma: (21) Terga shiny metallic 
green with central brown band (varies greatly in size) devoid or nearly 
devoid of metallic tints; very fine, inconspicuous basal band of tomentum; 
medium length white to yellowish hairs ventro-laterally on terga 2-4 and 
anteriorly on tergum 1; long black hairs dorsally on terga 5-6. (22) Sterna 
amber to brown (faint metallic tints on specimens from New Providence I.) ; 
long white to yellowish, scattered hairs on exposed areas. 

Agapostemon virescens (Fabricius) 

Andrena virescens Fabricius 1775. Type $ , British Museum (Natural History). 

Andrena nigricornis Fabricius 1793. Types £ , Zoologiske Museum, Copenhagen. 

Halictus dimidiatus Lepeletier 1841. Type $ , whereabouts unknown. 

Halictus tricolor Lepeletier 1841 (new synonymy). Type £, Institute e Museo di Zoologia, 

Universita di Torino, Italy. 
Agapostemon bicolor Robertson 1893. Type 2 , Illinois Natural History Survey. 

The synonymy of Andrena nigricornis Fabricius and Halictus dimidiatus 
Lepeletier with Andrena virescens Fabricius is based on their descriptions. 
C. D. Michener examined the type of A. virescens in the British Museum. 
G. C. Eickwort examined the type of Agapostemon bicolor in Urbana, 
Illinois, and confirmed Robertson's synonymy (1895) of this species with 
A. virescens. 1 have examined the type of hi. tricolor and it is a synonym 
of A. virescens, not A. radiatus as stated by Sandhouse (1936). 

Van der Vecht (1959) suggests that Apis sericea Forster (1771) may be 
a senior synonym of A. virescens, but the description given by Forster does 
not warrant such a conclusion (cf . Nomina Dubia) . 

Distribution (Fig. 30). Specimens have been collected as far east as 
Penobscot Co. in central Maine; as far west as Coos Co. on the Oregon 
Coast; as far south as Hernando Co. on the central western coast of Florida; 
and as far north as Vernon, British Columbia. This distribution is unusual 
inasmuch as it is wide, ranging from coast to coast, yet this species is com- 



548 The University Science Bulletin 

mon only from about 37° to 47° North latitude. In the northwestern part 
of the range (British Columbia and Alberta) males have been collected from 
August through October, and females from April through August and in 
October. In the central part of the range (Nebraska) males have been 
collected from July through October, and females from May through October. 
In the southeastern part of the range (Virginia, Tennessee, North Carolina 
and Georgia) males have been collected from June through September and 
in November, and females from May through August and in October. 
Although A. virescens occurs primarily at elevations of less than 2,000 ft. 
(610 m), I have seen specimens from as high as 4,000 ft. (1,219 m) on 
Middle Mountain, Virginia; 6,900 ft. (2,103 m) near Boulder, Colorado; and 
7,000 ft. (2,134 m) near Salt Lake City, Utah. 

Diagnosis. The male may be distinguished from many other North 
American species by the lack of yellow maculations on its brown to brown- 
black sterna 5-6, by its medial ridge on the apical ! /2 of sternum 6 (last 
visible sternum), and by its gonostylus (Fig. 183); from A. coloradinus and 
A. tyleri by the restriction of brown to brown-black areas on the posterior 
surface of its hind leg to basal and apical regions of femur and tibia, and 
by its lack of a prominent dorsal crest on its penis valve. The females may 
be distinguished from other species occurring north of Mexico by its black 
metasomal terga and the lack of yellow on its clypeus; from females of A. 
coloradinus and A. tyleri by the coarser carinae on its gena (2-3 per 0.25 
mm in A. virescens, 5-6 per 0.25 mm in A. coloradinus and A. tyleri), and 
its largely yellow mandibles (amber to brown in A. coloradinus and dark 
brown in A. tyleri). 

Description 
male (Figs. 97-98, 158, 183, 197) 
General coloration of head and mesosoma bright metallic green, meta- 
soma with black and yellow bands. Head (Figs. 97-98): pubescence white 
becoming yellowish to pale amber on vertex. (1) Labrum as in A. texanus 
but slightly blunter at apex. (2) Clypeus with medium size, shallow, 
scattered punctures on yellow portion, becoming more dense and slightly 
larger above. (3) Interocular area with large deep contiguous punctures 
below ocelli, becoming rugose and more shallowly sculptured below; supra- 
clypeal protuberance shallowly rugose to punctate with few faint horizontal 
rugae. (4) Vertex with deep contiguous punctures anteriorly, becoming 
more finely punctate laterally, and coarsely but shallowly rugose posteriorly 
and between ocelli. (5) Gena with numerous short, parallel, subcontiguous 
carinae extending dorso-laterally from antero-ventral margin. (6) Malar 
area bright yellow to pale amber; short. (7) Mandible bright yellow with 
apical V3 ferruginous. (8) Antenna brown to brown-black above with shiny 
pale area on apical l / 2 of apical flagellomere; flagellum pale amber to brown 



Revision of the Bee Genus Agapostemon 549 

below; and underside of scape and pedicel yellow to largely black with yellow 
apically. Mesosoma: pubescence white, becoming pale amber to yel- 
lowish on mesonotum and metanotum. (9) Pronotum with lateral angle 
and posterior lobe neither very rounded nor acute; postero-lateral area with 
large, broad, low, vertical ridge just anterior to numerous very fine and 
faint carinulae. (10) Mesoscuhim with fine contiguous punctures becoming 
finely rugose anteriorly and antero-laterally. (11) Mesoscutellum with punc- 
tures like those of mesoscutum but somewhat larger and less dense (especially 
laterally). (12) Metanotum finely and deeply rugose. (13) Mesepisternum 
coarsely rugose anteriorly, becoming finely rugose posteriorly. (14) Metepi- 
stcrnum with irregular horizontal carinae, anastomosing so as to appear 
largely rugose on some specimens. (15) Propodeum with distinct propodeal 
carina; propodeal shield coarsely and shallowly rugose; dorsal area coarsely 
rugose, postero-lateral area moderately rugose, becoming finely rugose to 
coarsely punctate antero-laterally. (16) Wing clear to transparent amber 
with apical regions slightly darkened. Radial vein dark brown, stigma and 
remaining veins pale brown. (17) Tegula pale transparent amber with 
green tints antero-basally, short transverse yellow band on anterior area 
curving rearward and almost reaching distal margin; posterior margin 
yellow. (18) Fore leg yellow with coxa bright metallic green; trochanter 
dark brown; femur with brown basally and postero-dorsally on basal x / 2 of 
tibia. Pubescence white to pale yellow. (19) Middle leg like fore leg but 
with only faint green tints on coxa and with brown streak postero-dorsally 
on basal %-% of tibia. Pubescence white to pale yellow. (20) Hind leg 
(Fig. 158) with coxa bright metallic green; trochanter brown (sometimes 
with faint metallic tints); femur yellow with brown at base and apex; 
tibia yellow with brown at base (and sometimes at apex) and brown streak 
centrally on antero-dorsal surface. Pubescence white to pale yellow. Meta- 
soma: (21) Terga brown to brown-black with yellow bands on basal 
halves of terga 2-5 and narrow yellow band centrally on tergum 1; pubes- 
cence white, short and appressed on dorsal portions of terga 1-4, moderate 
length elsewhere. (22) Sterna brown to brown-black with yellow baso- 
laterally on sterna 2-4 (often hidden by overlapping sterna), and sternum 1 
with metallic tints basally; apical ! / 2 of sternum 6 with median ridge in 
shallow depression. Pubescence white and moderately long on exposed 
areas. (23) Genitalia (Figs. 183, 197) with dorsal crest of penis valve 
rounded and broad, not prominent as in most other species; ventral lobes 
of moderate size and distally fringed with hairs. 

female (Figs. 28, 53-54) 

General coloration of head and mesosoma bright metallic green to 
blue-green, metasoma black with white hair bands. Head (Figs. 



550 The University Science Bulletin 

53-54) : pubescence white, becoming faintly yellowish at vertex (and on 
interocular area of some specimens). (1) Labium as in A. texanus. (2) 
Clypeus with large scattered punctures below, becoming contiguous above. 
(3) Interocular area with deep, medium sized punctures above, becoming 
deeply but finely rugose below. (4) Vertex with deep, small, contiguous 
punctures anteriorly, finer and shallower laterally and becoming finely rugose 
posteriorly; area between ocelli finely rugose to punctate. (5) Gena with 
moderately fine (2-3 per 0.25 mm) parallel carinae extending postero-dorsally 
from antero-ventral margin. (6) Malar area amber to ferruginous and very 
short, almost absent. (7) Mandible yellow or rarely very pale amber; with 
apical Yi- 1 /^ ferruginous. (8) Antenna dark brown to brown-black with 
underside of flagellum slightly paler brown. Mesosoma: pubescence white, 
becoming pale amber to slightly fuliginous on mesonotum and meta- 
notum. (9) Pronotum with lateral angle and posterior lobe neither very 
rounded nor acute; numerous very faint parallel carinulae extending antero- 
dorsally from postero-ventral area. (10) Mesoscutum with deep, medium 
sized, contiguous punctures becoming rugose anteriorly and antero-laterally. 
(11) Mesoscutellum with fine contiguous punctures medially, becoming 
large and separated by as much as 2-3 times their diameters laterally. (12) 
Metanotum rugulose. (13) Mesepisternum coarsely rugose, only very slightly 
finer rugae posteriorly. (14) Metepisternum with moderately coarse carinae 
regular and parallel anteriorly, irregular and anastomosing posteriorly. (15) 
Propodeum with prominent propodeal carina; propodeal shield with nu- 
merous weak carinae extending laterally and slightly dorsally from median 
groove; dorsal area with irregularly anastomosing carinae extending postero- 
laterally from anterior margin; lateral area with anastomosing horizontal 
rugae coarse posteriorly, becoming fine anteriorly. (16) Wing as in $ . 
(17) Tegula as in $ but slightly darker. (18-20) Fore, middle and hind legs 
brown to brown-black with yellow dorsally at apex of femur and base of tibia 
of fore and middle legs; dorsal surface of hind coxa with strong metallic 
tints; pubescence white to amber, becoming fuliginous apically on dorsal 
surface of middle leg and brown-black basally on antero-dorsal surface of 
hind leg. Metasoma: (21) Terga black; white tomentose hair bands 
basally on terga 2-4 and centrally on tergum 1 (often broadly interrupted 
medially); pubescence anteriorly on tergum 1 and laterally on terga 1-5 
white and moderately long; pubescence dorsally on terga 5-6 moderately long 
and brown to black. (22) Sterna brown to brown-black with long, scattered 
white hairs on exposed areas. 

Agapostemon viridulus (Fabricius) 

Apis viridula Fabricius 1793. Type $, Lund collection, Zoologiske Museum, Copenhagen. 
Indrena (Agapostemon) femoralis Guerin-Meneville 1844. Type $, Museo Civico di Storia 
Naturale di Genova, Genoa. 



Revision of the Bee Genus Agapostemon 551 

Agapostemon semiviridis Cresson 1865 (new synonymy). Type 9 , Academy of Natural Sciences, 

Philadelphia. 
Agapostemon stridulus. Dalla Torre, 1896. 

J. S. Moure has selected a lectotype of Apis viridula and redescribed it 
(Moure, 1960). C. D. Michener has examined the type of Andrena (Aga- 
postemon) jemoralis, and I have examined the lectotype of Agapostemon 
semiviridis. I have no doubt that all three are conspecific. It should be noted 
that owing to the inadequacy of the original description Agapostemon 
viridulus was believed by many authors to be a synonym of Agapostemon 
virescens. 

Except for the coloration of the head and thorax this species is indis- 
tinguishable from Agapostemon obscitratus Cresson. Sympatry and the 
absence of intermediate forms lead me to conclude that A. obscitratus is not 
conspecific with A. viridulus. 

Distribution. I have seen specimens from the following localities in 
Cuba: Baracoa (3 $ , Aug.), Castillo de Juaga, Cienfuegos (8 2 , 4-5 Sept.), 
Central Jaronu (4 8 , 31 Mar.), Cristo (4 8 , 3 Oct.), Guane (2 8 & 1 2 , 
24-26 Sept.), Guantanamo (5 8 & 1 2 ), Havana (8 $), Puerto Baniato, 
Santiago de Cuba (1 $ , Nov.), Rio Toa, Baracoa (2 8 , 26 Apr.), San Bias 
(1 8 , 12 Aug.), Santiago de las Vegas (1 9 , 10 Feb.), Siboney (12, Feb.), 
Soledad (1 8,27 Feb.; 1 8 , 4 Mar.; 1 8 , 5 May), Trinidad Mts. (1 2 , 
28 Aug.), Vinales (2 2 , 16-22 Sept.). 

Diagnosis. The male may be distinguished from West Indian species, 
except A. obscuratus and A. hispanioticus, by its black metasomal terga; 
from A. obscuratus by its bright metallic blue-green head and mesosoma; 
from A. hispaniolicus by the subcontiguous punctures of its mesoscutum. 
The female may be distinguished from West Indian species, except A. 
obscuratus, A. hispaniolicus and A. swainsonae, by its black metasomal 
terga; from A. obscuratus by its bright metallic blue-green head and meso- 
soma; from A. swainsonae by the metallic green tints on the base of its 
mandible; and from A. hispaniolicus by the coarser sculpturing of its in- 
terocular area. 

Description 
male (Figs. 111-112, 150, 178, 208) 
General coloration of head and mesosoma metallic green to blue-green, 
metasoma dark brown to black. Head (Figs. 111-112): (1) Labrum 
yellow to white on basal l / 2 , amber to brown on apical l / 2 ; apex smoothly 
rounded (not acute), basal l / 2 with shallow depression medially on trans- 
verse ridge; shiny and impunctate; pubescence limited to about 20 subapical 
bristles. (2) Clypeus yellow to white below and extraordinarily elongate; 
punctures coarse and scattered, weak transverse rugae medially on upper Y 2 ; 
pubescence short, sparse and white. (3) lnterocular area with contiguous 



552 The University Science Bulletin 

punctures below median ocellus, becoming rugose midway between median 
ocellus and antennal sockets, rugosity at level of antennal sockets changing 
to scattered punctures ventro-laterally; supraclypeal protuberance with coarse 
scattered punctures and transverse medial rugae; pubescence white, nearly 
as long between antennal sockets as scape, becoming shorter laterally, 
dorsally and ventrally; ventrolateral area near malar space with short, 
sparse, white tomentum. (4) Vertex with fine, contiguous punctures an- 
teriorly and laterally, becoming rugose posteriorly; area between ocelli 
sparsely punctate; pubescence white, fuliginous between ocelli. (5) Genu 
with fine parallel rugae running postero-dorsally from antero-ventral margin; 
pubescence white and about 1.5 times as long postero-medially as pubescence 
between antennal sockets, becoming shorter laterally, anteriorly and dorsally. 
(6) Malar area yellowish, more than l / z as broad as long; shiny and im- 
punctate; pubescence sparse, appressed, short and white. (7) Mandible 
yellow to creamy, apical y$- x A transparent ferruginous; simple hairs on 
postero-ventral margin of mandible longest basally, becoming y 5 as long 
apically. (8) Antenna with scape brown-black above, yellow below; pedicel 
dark brown; flagellum brown-black above, paler brown below. Meso- 
soma: pubescence fuliginous on mesonotum and metanotum, white 
laterally, ventrally and on propodeum. (9) Pronotum with lateral angle and 
posterior lobe rounded; sculpturing inconspicuous. Pubescence postero- 
laterally very fine, short and inconspicuous; dorsally on lateral angle and 
posterior lobe similar in length and color to that of mesonotum. (10) 
Mesoscutum with very fine, deep, contiguous punctures, becoming rugulose 
antero-laterally. (11) MesoscntelUim shiny with large, scattered, central 
punctures, becoming finer and more numerous anteriorly. (12) Metanotum 
finely and irregularly rugulose with large, scattered punctures. (13) Mesepi- 
sternum coarsely to moderately rugose anteriorly, becoming finely rugose 
posteriorly. (14) Metepisternum finely and horizontally carinulate. (15) 
Propodeum rounded posteriorly; propodeal carina absent; fine, even, 
horizontal carinulae laterally and posteriorly (uppermost of these becoming 
stronger along upper margin of propodeal shield) ; dorsal area of propodeum 
with fine, even carinulae running posten>laterally from posterior margin 
of metanotum. (16) Wing transparent, faintly fuliginous; veins and ptero- 
stigma dark brown, radius almost black. (17) Tegula dark transparent 
brown with metallic green tints on basal margin and submarginally on 
anterior ! / 3 and with very fine scattered punctures. (18) Fore leg dark 
brown with yellow to white on antero-ventral surface of femur, on all but 
posterior surface of tibia, and on tarsus; coxa metallic green; pubescence 
white proximally, becoming amber distally. (19) Middle leg similar in 
color and pubescence to fore leg but with white to yellow areas on femur 
and tibia reduced, and with tarsus amber to brown, coxa slightly metallic 



Revision of the Bee Genus Agapostemon 553 

green. (20) Hind leg (Fig. 150) with coxa largely metallic green to blue; 
trochanter brown with metallic tints; femur brown with yellow to white 
antero-ventrally, on basal % antero-dorsally, and posteriorly around large 
tooth; tibia brown with yellow to white ventrally and antero-ventrally 
(yellow sometimes very reduced or absent); tarsus brown to amber. Hind 
leg very swollen, especially femur; basal ridge on basitarsus very prominent, 
broad and concave; apical groove broad and delimited by strong carina; 
pubescence amber on tarsus, white on other segments. Metasoma: (21) 
Terga brown-black (pygidium somewhat paler) and shiny; with long, 
sparse, white pubescence becoming fuliginous around pygidium and with 
very short, moderately dense pubescence brown-black and inconspicuous, 
and with narrow, basal bands of white tomentum on terga 2-6 (often hidden 
by preceding terga). (22) Sterna not quite as dark as terga, sternum 1 with 
metallic green tints basally; with sparse white pubescence on exposed por- 
tions. (23) Genitalia (Figs. 178, 208) with penis valves smooth basally on 
dorsal surface, with large dorsal crest grooved at apex. Gonocoxa with 
numerous deep longitudinal grooves laterally; gonostylus with simple apical 
stylus; medial plate with small fold on antero-medial margin; ventral lobes 
large, curled posteriorly and with dense, short, simple pubescence marginally 
on posterior surface. 

female (Figs. 33-34) 

General coloration of head and mesosoma metallic green to blue-green, 
metasoma black with narrow bands of white tomentum. Head (Figs. 33- 
34) : pubescence white, fuliginous on vertex, and amber on lower margin of 
clypeus and on mandibles. (1) Labrum (Figs. 34 A-C) brown, with large, 
broad, distal keel; basal area with rounded medial protuberance and sharply 
delimited from distal process by marginal carina; additional transverse 
carina present between base of distal process and outer margin of basal area. 
(2) Clypeus elongate and with broad, very shallow, subapical depression or 
flattened area; punctures very coarse, widely separated distally, becoming 
more numerous above; irregular transverse rugae often scattered on metallic 
area. (3) Interocular area rugose, becoming punctate at vertex; supraclypeal 
protuberance with scattered punctures and regular transverse rugae. (4) 
Vertex finely and densely punctate anterior to and lateral to ocellar triangle, 
becoming shallowly and weakly rugose posterior to ocelli; area between 
ocelli with fine, scattered punctures. (5) Gena with numerous, regular, con- 
tiguous rugulae running postero-dorsally from antero-ventral margin. (6) 
Malar area very short, almost non-existent. (7) Mandible dark ferruginous 
brown with metallic green tints at base. (8) Antenna brown-black, slightly 
paler on underside of flagellum. Mesosoma: pubescence as in 
Pronotum with lateral angle and posterior lobe rounded; weak carina 
running postero-laterally downward from lateral angle. (10) Mesoscutum 



554 The University Science Bulletin 

with dense, fine punctures contiguous centrally, becoming separated by 
about l / 2 tbeir diameters posteriorly and becoming finely rugulose anteriorly. 
(11) Mesoscutellum with scattered medium punctures interspersed with 
more numerous fine punctures anteriorly, and finely rugulose posteriorly and 
laterally. (12) Metanotum with very faint rugulae running postero-laterally 
from anterior margin. (13) Mesepisternum moderately rugose; rugae 
moderately prominent anteriorly, becoming finer posteriorly. (14) Metepi- 
sternum finely, horizontally carinulate, with carinulae more numerous and 
less prominent than on S . (15) Propodeum not rounded posteriorly and 
with definite propodeal carina; carinulae on lateral and dorsal areas similar 
to $ ; numerous parallel carinulae on propodeal shield running dorso- 
lateral^ from median groove. (16-17) Wing and tegula as on $ . (18-20) 
Fore, middle and hind legs brown-black to amber; pubescence creamy to 
amber, fuliginous on dorsal surfaces of tibiae, dark brown to black adjacent to 
basi-tibial plate on hind leg; posterior tibial spur with 3-5 (usually 4) 
spatulate teeth. Metasoma: (21) Terga brown-black to black, anterior 
area of tergum 1 often brown to amber; dorsal surfaces finely granulose, 
not shiny; pubescence on ventrolateral areas of terga 1-4 as well as on 
anterior surface of tergum 1 scattered, of moderate length and white; dorsal 
portions of terga with very numerous, minute, simple, brown-black hairs 
directed posteriorly; basal bands of white tomentum on terga 2-5 usually 
hidden by preceding terga but may be visible laterally; abundant long 
brown-black pubescence on terga 5-6. (22) Sterna brown to black with 
scattered, simple, long white hairs on exposed portions. 

NOMINA DUBIA 

In 1771, J. R. Forster described Apis sericea from North America. Van 
der Vecht (1959) has suggested that it is a senior synonym of Agapostemon 
virescens Fabricius 1775. Forster 's very brief description is sufficient only 
to indicate that he had a male Agapostemon. Forster made no mention of 
characters of specific significance and, in the absence of the type, I can 
accord only the status of nomen dubium to Apis sericea Forster. 

In 1903, J. Vachal described Ha/ictus (Agapostemon) chiriquiensis from 
a single female. It would seem from the description to be either A. nasutus 
or A. leuncidus. I requested the type but Dr. S. Kelner-Pillault was unable 
to find it in the Museum National D'Histoire Naturelle in Paris. Therefore, 
I consider Halictus (Agapostemon) chiriquiensis a nomen dubium. 



Revision of the Bee Genus Agapostemon 555 

APPENDIX A 

The following list of plant communities in which Agapostemon texanus and A. angelicus 
may occur was derived by superimposing distributions of these species on the vegetation map 
of Kiichler (1964). Only those communities present in the United States are listed, as com- 
parable vegetation maps do not exist for Mexico and Central America. The letter "T" preceding 
a community indicates records of A. texanus, "A" indicates A. angelicus, and "TA" indicates 
both species. The numbers are those assigned to the communities by Kiichler. 

As Kiichler's map is based on potential vegetation, this list may be unrealistically large, 
with the bees actually living in disturbed areas with little, if any, semblance of the potential 
climax or subclimax vegetation. 

T 1. Spruce — Cedar — Hemlock Forest (Picea — Thuja — Tsuga) 

T 2. Cedar — Hemlock. — Douglas Fir Forest (Thuja — Tsuga — Pseudotsuga) 

T 3. Silver Fir — Douglas Fir Forest (Abies — Pseudotsuga) 

T 5. Mixed Conifer Forest (Abies — Pinus — Pseudotsuga) 

T 6. Redwood Forest (Sequoia — Pseudotsuga) 

T 7. Red Fir Forest (Abies) 

T 8. Lodgepole Pine — Subalpine Forest (Pinus — Tsuga) 

T 9. Pine — Cypress Forest (Pinus — Cypressus) 

T 10. Ponderosa Shrub Forest (Pinus) 

T 11. Western Ponderosa Forest (Pinus) 

T 12. Douglas Fir Forest (Pseudotsuga) 

T 13. Cedar — Hemlock — Pine Forest (Thuja — Tsuga — Pinus) 

T 14. Grand Fir — Douglas Fir Forest (Abies — Pseudotsuga) 

A 15. Western Spruce — Fir Forest (Picea — Abies) 

TA 17. Black Hills Pine Forest (Pinus) 

TA 18. Pine — Douglas Fir Forest ( Pinus — Pseudotsuga) 

A 19. Arizona Pine Forest (Pinus) 

A 20. Spruce — Fir — Douglas Fir Forest (Picea — Abies — Pseudotsuga) 

A 21. Southwestern Spruce — Fir Forest (Picea — Abies) 

TA 23. Juniper — Pinyon Woodland (Juniperus — Pinus) 

T 29. California Mixed Evergreen Forest (Quercus — Arbutus — Pseudotsuga) 

T 30. California Oakwoods (Ouei-cus) 

A 31. Oak — Juniper Woodland (Quercus — Juniperus) 

TA 32. Transition between 31 and 37 

TA 33. Chaparral (Adenostoma — Arctostaphylos — Ceanothus) 

T 35. Coastal Sagebrush (Salvia — Eriogonum) 

T 36. Mosaic of numbers 30 and 35 

TA 37. Mountain Mahogany — -Oak Scrub (Cercocarpus — Quercus) 

TA 38. Great Basin Sagebrush (Artemesia) 

A 39. Blackbrush (Coleogyne) 

A 40. Saltbush — Greasewood (Atriplex — Sarcobatus) 

TA 41. Creosote Bush (Larrea) 

TA 42. Creosote Bush — Bur Sage (Larrea — Franseria) 

TA 43. Palo Verde — Cactus Shrub (Cercidium — Opuntia) 

TA 44. Creosote Bush — Tarbush (Larrea — Flourensia) 

TA 45. Ceniza Shrub (Leucophyllum — Larrea — Prosopis) 

T 47. Fescue — Oatgrass (Festuca — Danthonia) 

T 48. California Steppe (Stipa) 

T 50. Fescue — Wheatgrass (Festuca — Agropyron) 

T 51. Wheatgrass — Blue Grass (Agropyron — Poa) 

TA 53. Grama — Galleta Steppe (Bouteloua — Hilaria) 

A 54. Grama — Tobosa Prairie (Bouteloua — Hilaria) 

TA 55. Sagebrush Steppe (Artemesia — Agropyron) 

A 58. Grama— Tobosa Shrub Steppe (Bouteloua— Hilaria — Larrea) 

TA 59. Trans-Pecos Shrub Savanna (Flourensia — Larrea) 

TA 60. Mesquite Savanna (Prosopis — Hilaria) 



556 The University Science Bulletin 

T 61. Mesquite — Acacia Savanna (Prosopis — Acacia — Andropogon — Setaria) 

T 62. Mesquite — Live Oak Savanna (Prosopis — Quercus — Andropogon) 

T 63. Foothills Prairie (Agropyron — Festuca — Stipa) 

T 64. Grama — Needlegrass — Wheatgrass (Boitteloua — Stipa — Agropyron) 

TA 65. Grama — Buffalo Grass (Bonteloua — Buchlo'e) 

TA 66. Wheatgrass — Needlegrass (Agropyron — Stipa) 

TA 67. Wheatgrass — Bluestem — Needlegrass (Agropyron — Andropogon — Stipa) 

TA 68. Wheatgrass — Grama — Buffalo Grass (Agropyron — Botitelotta — Buchlo'e) 

TA 69. Blue-stem— Grama Prairie (Andropogon — Boutcloua) 

A 70. Sandsage — Bluestem Prairie (Artemesia — Andropogon) 

TA 71. Shinnery (Quercus — Andropogon) 

T 74. Bluestem Prairie (Andropogon — Panicum — Sorghastrum) 

TA 75. Nebraska Sandhills Prairie (Andropogon — Calamovilja) 

TA 76. Blackland Prairie (Andropogon — Stipa) 

T 77. Bluestem — Sacahuista Prairie (Andropogon — Spartina) 

T 81. Oak Savanna (Quercus — Andro'pogon) 

T 82. Mosaic of numbers 74 and 100 

T 84. Cross Timbers (Quercus — Andropogon) 

TA 85. Mesquite — Buffalo Grass (Prosopis — Buchlo'e) 

TA 86. Juniper — Oak Savanna (Junipcrus — Quercus — Andropogon ) 

TA 87. Mesquite — Oak Savanna (Prosopis — Quercus — Andropogon) 

T 88. Fayette Prairie (Andropogon — Buchlo'e) 

T 95. Great Lakes Pine Forest (Pinus) 

TA 98. Northern Floodplain Forest (Populus — Salix — Ulnius) 

T 99. Maple — Basswood Forest (Acer — Tilia) 

T 100. Oak — Hickory Forest (Quercus — Carya) 

T 102. Beech — Maple Forest (Fagus — Acer) 

T 104. Appalachian Oak Forest (Quercus) 

T 106. Northern Hardwoods (Acer — Betula — Fagus — Tsuga) 

T 109. Transition between numbers 105 and 106 

T 110. Northeastern Oak — Pine Forest (Quercus — Pinus) 

T 111. Oak — Hickory — Pine Forest (Quercus — Carya — Pinus) 

T 113. Southern Floodplain Forest (Quercus — Nyssa — Taxodium) 



LITERATURE CITED 

(Anonymous). 1964. International code of zoological nomenclature. London: International 

Trust for Zoological Nomenclature. 176 pp. 
\siimead, W. H. 1890. On the Hymenoptera of Colorado; descriptions of new species, notes, 

and a list of the species found in the state. Colorado Biol. Assoc. Bull. 1:408-409. 
Cameron, P. 1902. Descriptions of new genera and species of American Hymenoptera. Trans. 
Amer. Entomol. Soc. 28:369-377. 
— . 1903. Descriptions of new species of Hymenoptera taken by Mr. Edward Whymper 
on the "Higher Andes of the Equator." Trans. Amer. Entomol. Soc. 29:225-238. 
Cockereli., T. D. A. 1898. Note on Agapostemon texanus. Entomol. News 9:27. 

— . 1900. Descriptions of new bees collected by Mr. H. H. Smith in Brazil — I. Proc. Acad. 

Natur. Sci. Philadelphia, pp. 356-377. 
— . 1905. Notes on some bees in the British Museum. Trans. Amer. Entomol. Soc. 

31:309-364. 
— . 1909. Descriptions and records of bees-XXI. Ann. Mag. Natur. Hist. (8)4:25-31. 
— . 1910a. The north American bees of the genus Nomia. Proc. U.S. Nat. Mus. 
38:289-298. 

. 1910b. Some neotropical bees. Psyche 17:142-143. 

. 1910c. Descriptions and records of bees-XXVII. Ann. Mag. Natur. Hist. (8)5:361-369. 
— . I'M 2a. Descriptions and records of bees-XLIV. Ann. Mag. Natur. Hist. (8)9:554-568. 
— . 1912b. Descriptions and records of bees-XLV. Ann. Mag. Natur. Hist. (8)10:21-31. 
— . 1917a. Descriptions and records of bees-LXXVI. Ann. Mag. Natur. Hist. (8)20:235-241. 
— . 1917b. Descriptions and records of bees-LXXVIII. Ann. Mag. Natur. Hist. (8)20:436- 
441. 



Revision of the Bee Genus Agapostemon 557 

. 1918a. New halictine bees from Chile. Canadian Entomol. 50:343-345. 

. 1918b. Descriptions and records of bees-LXXXI. Ann. Mag. Natur. Hist. (9)2:418-425. 

. 1919. Bees in the collection of the United States National Museum-3. Proc. U.S. Nat. 

Mus. 55:167-221. 
. 1924. Expedition of the California Academy of Sciences to the Gulf of California in 

1921. Proc. California Acad. Sci. 12:529-560.' 
. 1925. Bees in the collection of the California Academy of Sciences. Proc. California 

Acad. Sci. 14:185-215. 
. 1927. Bees of the genera Agapostemon and Augochlora in the collection of the 

California Academy of Sciences. Pan-Pacific Entomol. 3:153-162. 

. 1937a. New and little-known American bees. Amer. Mus. Novitates No. 889, pp. 1-5. 

. 1937b. Bees from San Miguel Island, California. Pan-Pacific Entomol. 13:148-158. 

■ . 1939. The bees of southern California Islands. Proc. California Acad. Sci. 23:427-436. 

. 1949. Bees from Central America, principally Honduras. Proc. U.S. Nat. Mus. 

98(3233) :429-490. 
Crawford, J. C, Jr. 1901. North American bees of the genus Agapostemon Guerin. Proc. 

Nebraska Acad. Sci. 7:156-165. 

■. 1906. Some Costa Rican bees. Trans. Amer. Entomol. Soc. 32:157-163. 

Cresson, E. T. 1865. On the Hymenoptera of Cuba. Proc. Entomol. Soc. Philadelphia 1865: 

1-200. 
. 1869. Notes on Cuban Hymenoptera with descriptions of new species. Trans. Amer. 

Entomol. Soc. 2:293-298. 

. 1872. Hymenoptera Texana. Trans. Amer. Entomol. Soc. 4:153-292. 

. 1874. Descriptions of new Hymenoptera. Trans. Amer. Entomol. Soc. 5:99-102. 

. 1875. Report upon the collections of Hymenoptera made in portions of Nevada, Utah, 

Colorado, New Mexico, and Arizona, during the years 1872, 1873, and 1874. Report 

U.S. Geogr. Survey 5:705-736. 
Dalla Torre, C. G. 1896. Catalogus Hymenopterorum. Leipzig. 10:643 pp. 
Eickwort, G. C, and K. R. Eickwort. 1969. Aspects of the biology of Costa Rican halictine 

bees, I. Agapostemon nasutus (Hymenoptera: Halictidae). J. Kansas Entomol. Soc. 

42(4):421-452. 
Fabricius, J. C. 1775. Systema Entomologiae. Flensburgi et Lipsiae. 832 pp. 
. 1793. Entomologia Systematica Emendata et Aucta. Copenhagen. Christ. Gottl. Proft. 

2:519 pp. 
Forster, J. R. 1771. Novae Species Insectorum, Centuria I. London. 100 pp. 
Friese, H. 1916. Zur bienenfauna von Costa Rica (Hym.). Stett. Entomol. Zeit. 77:287-348. 
Guerin-Meneville, F. E. 1844. Iconographie du Regne Animal de G. Cuvier, Insectes. 

3:447-448. 
Holmberg, E. L. 1903. Delectus Hvmenopterologicus Argentinus. Anales del Museo Nacional 

de Buenos Aires. (3)2:377-517. 
Kuchler, A. W. 1964. Manual to Accompany the Map Potential Natural Vegetation of the 

Conterminous United States. Amer. Geogr. Soc. Special Publ. 36. 116 pp. 
Lepeletier, L. M. 1841. Histoire Naturelle des Insectes; Hymenopteres. Paris. 2:680 pp. 
Lucas, P. H. 1856. Ordre des hymenopteres. In Ramon de la Sagra, Histoire physique, 

politique et naturelle de l'ile de Cuba. (Insects) Paris. 7:775-776. 
Michener, C. D. 1937. Records and descriptions of North American bees. Ann. Mag. Natur. 

Hist. (10)19:313-410. 
. 1944. Comparative external morphology, phylogeny, and a classification of the bees 

(Hymenoptera). Bull. Amer. Mus. Natur. Hist. 82:151-326. 

. 1954. Bees of Panama. Bull. Amer. Mus. Natur. Hist. 104:1-175. 

. 1965. A classification of the bees of the Australian and South Pacific regions. Bull. 

Amer. Mus. Natur. Hist. 130:5-362. 
Michener, C. D., and R. B. Lange. 1958. Observations on the behavior of Brazilian halictid 

bees (Hymenoptera, Apoidea) I. Ann. Entomol. Soc. Amer. 51:155-164. 
Mitchell, T. B. 1960. Bees of the Eastern United States. Vol. 1. North Carolina Agr. Exp. 

Station Tech. Bull. No. 141, 538 pp. 
Moure, J. S. 1940. Apoidea Neotropica. Rev. Mus. Paulista (25)2:39-64. 
. 1947. Sobre algunas abejas de la provincia Salta. Rev. Soc. Entomol. Argentina 

13:218-253. 
. 1960. Notes on the types of the neotropical bees described by Fabricius. Studia Entomol. 

[Rio] 3:103-104. 
. 1964. Two new genera of halictine bees from the Araucanian subregion of South 

America (Hymenoptera; Apoidea). J. Kansas Entomol. Soc. 37:265-275. 



558 The University Science Bulletin 

Muesebeck, C. F. W., K. V. Krombein, H. K. Townes, et al. 1951. Hymenoptera of America 

North of Mexico. Svnoptic Catalog. U.S.D.A. Agr. Monogr. No. 2, Washington, D.C. 

1420 pp. 
Ribble, D. W. 1965. A revision of the banded subgenera of Nomia in America (Hvmenoptera: 

Hahctidae). Univ. Kansas Sci. Bull. 45(3) :277-359. 
Ridgway, R. 1912. Color Standards and Color Nomenclature. A. Hoen Co., Washington, D.C. 

47 pp. 
Roberts, R. B. 1969. Biologv of the bee genus Agapostemon (Hymenoptera: Halictidae). 

Univ. Kansas Sci. Bull. 48(16) :689-719. 
Robertson, C. 1893. Notes on bees, with descriptions of new species. Trans. Amer. Entomol. 

Soc. 20:145-149. 
. 1895. Notes on bees, with descriptions of new species. Trans. Amer. Entomol. Soc. 

22:115-128. 
. 1897. North American bees-descriptions and synonyms. Trans. Acad. Sci. St. Louis 

7:315-356. 

. 1902. Synopsis of Halictinae. Canadian Entomol. 34:243-250. 

Say, T. 1837. Descriptions of new species of North American Hymenoptera and observations 

on some already described. Boston J. Natur. Hist. 1 (4) :361-416. 
Sandhouse, G. A. 1936. The bees of the genus Agapostemon (Hymenoptera: Apoidea) oc- 
curring in the United States. J. Washington Acad. Sci. 26:70-83. 
Schrottky, C. 1902a. Hymenopteres nouveaux de l'Amerique meridionale. Anales Mus. 

Nac. Buenos Aires 7:309-315. 

. 1902b. Ensaio sobre as abelhas solitarias do Brazil. Rev. Mus. Paulista 5:330-613. 

. 1903. Enumeration des Hymenopteres connus jusqu'ici de la Republique Argentine, de 

l'Uruguay et du Paraguay. Anales Soc. Cient. Argentina 55:176-186. 

. 1908. Nuevos Himenopteros. Anales Soc. Cient. Argentina 65:225-239. 

.1909a. Nuevos Himenopteros Sudamericanos. Rev. Mus. La Plata (2)16:137-149. 

. 1909b. Synonymische bemerkungen iiber einige sud-amerikanische Halictinae (Hym.). 

Deutsche Entomol. Zeitschr. 4:479-485. 
. 1913. Distribucion Geographica de los Himenopteros Argentinos. Anales de la 

Sociedad Cientifica Argentina. 75(5) :240. 
Smith, F. 1853. Catalogue of Hymenopterous Insects in the Collection of the British Museum. 

Part 1. Cat. Hym. British Mus. 1:197 pp. 
. 1879. Descriptions of New Species of Hymenoptera in the Collection of the British 

Museum. London, 240 pp. 
Spinola, M. 1851. Himenopteros. I. Apisiteos. II. Andrenitas. In Claudio Gay, Hist. Fis. 

Pol. Chile, Zoologia 6:158-232. 
Strand, E. 1910. Beitrage zur kenntnis der hymenopteren fauna von Paraguay. Zool. 

Jahrbucher, Abteilung Syst. Geog. Biol. Tiere 29:455-562. 
Tins, E. S. G. 1901. Notes on Colorado bees. Canadian Entomol. 32:303-305. 
Vachal, J. 1901. Contributions hymenopteriques. Hymenoptera mellifera americana nova. 

Ann. Soc. Entomol. France 68:77-81. 
— . 1903. Etude sur les Halictus d'Amerique (Hym.). Misc. Entomol. 11:89-104, 121-136. 
—.1904. Etude sur les Halictus d'Amerique (Hym.). Misc. Entomol. 12:9-24, 113-128, 

137-144. 
Vecht, J. van der. 1959. Notes on aculeate Hymenoptera described in the period 1758-1810. 

Entomol. Berichten. 19:65-70. 
\\ estwood, J. O. 1875. Descriptions of some new species of short-tongued bees belonging to 

the genus Nomia of Latreille. Trans. Entomol. Soc. London (3)1875:207-222. 
Wolcott, G. N. 1948. Insects of Puerto Rico: Apoidea. J. Agr. Univ. Puerto Rico 32(4): 

865-873. 



Revision of the Bee Genus Agapostemon 



559 




Fig. 1. Distributions of West Indian Agapostemon (stippled regions less than 100 fathoms 
in depth). 



-a 
c 
< 



c 
o 



rt 
U 



a 
o 

s ™ 

^ 



3 
OS 



bo 
c 
o 
l-J 



u 

-^ 

o 
o 

IH 

u 



rt 

c 
rt 

bo 
rt 



rt 
3 

u 



rt 






u 




rt 


O 






PS 
O 




u 


C 


rt 
u 




u 


Pi 


rt 




rt 


3 






rt 


C 
O 


Li 
3 
P-. 


a" 

> 


O 

U 



poeyi 

I'iequesensis 

ochromops 

columbi 

sapphirinus 

cyaneus 

insularis 

jamaicensis 

aenigma 



X 



X 
X 
X 



X 
X 
X 



X 



X 



X 



X 



X 



X 



X X 



XXX 



X 



X 



viridulus 

obscaratas 
cubensis 
alayoi 
hispaniolicus 

swainsonae 



X 
X 
X 



X 
X 



X 



\ohliellns 



XXX 



centratus 



X 



560 



The University Science Bulletin 




3= 


z (f 


€= 


-9 


• 


=o"&? 


A. cockerelli 



Fig. 2. Distribution of Agapostemon cockerelli. 



Revision of the Bee Genus Agapostemon 



561 




Fig. 3. Distribution of Agapostemon coloradimts. 



562 



The University Science Bulletin 




Fig. 4. Distribution of Agapostemon jemoratus. 



Revision of the Bee Genus Agapostemon 



563 





jf^ 








\ 5 




'••€.! 












1/ 






.-' 


.' 


J 3 


3=d" 

C=9 
•=d"&9 


s/VS 


■^-A,--' 








A. leuncul 


us 






) $ 



Fig. 5. Distribution of Agapostemon leuncidus. 



564 



The University Science Bulletin 




= 9 pole al 

tf^ = Qdark abdomen 

C) = 9 P ale & dark abd - 
^fc = 0T g, O pale abdomen 

^ —Qf 1 & y dark abdomen 

^fc=QT T &YP a l e & dark abdomen 

A. melliventris 



© 



3 



Fig. 6. Distribution of Agapostemon mellivetitns. Color of the metasomal terga of females 
is indicated by "dark abdomen" (= brown-black to black) and "pale abdomen" (= amber). 



Revision of the Bee Genus Agapostemon 



565 




Fig. 7. Distribution of Agapostemon nasutus with arrows indicating localities where females 
with pale amber metasomal terga have been found. 



566 



The University Science Bulletin 




Figs. 8-13. Coarse sculpturing of A. poeyt (left) and finer sculpturing of A. viequesensis 
(right): 8-9, mesonotum; 10-11, propodeal shield; 12-13, dorsal region of propodeum. 



Revision of the Bee Genus Agapostemon 



567 




Figs. 14-17. Coarse sculpturing of A. poeyi (left) and finer sculpturing of A. vieqtiesensis 
(right): 14-15, lateral view of mesosoma; 16-17, ventral view of genal region. 



568 



The University Science Bulletin 




Fig. 18. Distribution of Agapostemon radiattts. 



Revision of the Bee Genus Agapostemon 



569 




Fig. 19. Distribution of Agapostemon splendens. 



570 



The University Science Bulletin 




Fig. 20. Map showing distribution of Agapostemon texanus males and females (vertically 
divided circles); Agapostemon angelicas males only (horizontally divided circles); A. texanus 
and A. angelicas males (black circles); A. texanus and/or A. angelicas females (open circles). 



Revision of the Bee Genus Agapostemon 



571 




Figs. 
cence re 
turing. 



21-26. Agapostemon texanus females: 21-22, head with pubescence, and with F 
moved; 23-24, variation in mesoscutal punctation; 25-26, variation in propodeal sculp- 
Note propodeal carina (Fig. 25, arrow) characteristic of female Agaposten, 



572 



The University Science Bulletin 




Fig. 27. Distribution of Agapostemon tyleri. 



Revision of the Bee Genus Agapostemon 



573 




Figs. 28-29. Ventral view of genal region of Agapostemon virescens (left) and A. tyleri (right). 




Fig. 30. Distribution of Agapostemon virescens, 



574 



The University Science Bulletin 




Figs. 31-46. Heads of female Agapostemon in anterior and lateral views: 31-32, A. 
\ohliellus (stippling yellow) with lateral (A) and anterior (B) views of labrum; 33-34, A. 
viridulus (stippling brown) with anterior (A), antero-distal (B) and lateral (C) views of 
labrum; 35-36, A. swainsonae (stippling amber); 37-38, A. hispaniolicus (stippling brown) with 
anterior view (A) of apex of labrum; 39-40, A. vieqaesensis (stippling brown); 41-42, A. 
nasutus (stippling yellow); 43-44, A. atrocaeruleus; 45-46, A. leunculus (stippling yellow). 



Revision of the Bee Genus Agapostemon 



575 




Figs. 47-62. Heads of female Agapostemon in anterior and lateral views: 47-48, A. inter- 
medins (stippling yellow); 49-50, A. coloradinns (stippling brown-black); 51-52, A. tyleri 
(stippling brown-black); 53-54, A. virescens (stippling brown-black, hatching yellow); 55-56, 
A. radiatus (stippling black, hatching yellow); 57-58, A. femoratus (stippling black, hatching 
orange); 59-60, A. coc\erelli (stippling black, hatching yellow); 61-62, A. melliventris (stippl- 
ing yellow). 



576 



The University Science Bulletin 




Figs. 63-70. Heads of female Agapostemon in anterior and lateral views: 63-64, A, 
mexicanus (stippling amber, hatching yellow); 65-66, A. peninsularis (stippling yellow); 67-68, 
A. texanus or A. angelicus (stippling black, hatching yellow); 69-70, A. splendens (stippling 
brown-black, hatching amber). 

Figs. 71-78. Heads of male Agapostemon in anterior and lateral views (stippling yellow): 
71-72, ./. splendens; 73-74, A. angelicus; 75-76, A. texanus; 77-78, A. rhopalocera. 



Revision of the Bee Genus Agapostemon 



577 




Figs. 79-94. Heads of male Agapostemon in anterior and lateral views (Fig. 86, ventral 
view): 79-80, A. atrocaeruleus ; 81-82, A. intermedins; 83-84, A. leunculus; 85-87, A. nasutus; 
88-89, A. peninsnlaris; 90-91, A. mexicanus; 92-94, A. mellirentris (note size variation of two 
specimens, 93-94, collected at the same time and place). Yellow or creamy maculations are 
indicated with stippling. 



578 



The University Science Bulletin 




Figs. 95-108. Heads of male Agapostemon in anterior and lateral views: 95-96, A. tyleri; 
97-98, A. virescens; 99-100, A. coloradimts; 101-104, .-/. jemoratus (note size variation, 101- 
102); 105-106, A. radiatus; 107-108, A. coc/^erelli. Yellow or creamy maculations 
indicated with stippling. 



are 



Revision of the Bee Genus Agapostemon 



579 




Figs. 109-124. Heads of male Agapostemon in anterior and lateral views: 109-110, A. 
\ohliellus; 111-112, A. viridulus, 113-114, A. swainsonae; 115-116, A. hispaniolicus; 117-118, 
A. viequesensis; 119-120, A. cubensis; 121-122, A. centratus; 123-124, A. semimelleus. Yellow 
or creamy maculations are indicated with stippling. 



580 



The University Science Bulletin 





133 



136 



137 



138 







135 



140 




125-135 



136-140 



Figs. 125-140. 125-130, heads of Agapostemon in anterior and lateral views, with yellow 
maculations stippled: 125-126, A. mourei female; 127-128, A. mourei male; 129-130, A. semi- 
melleus female; 131, dorsal view of A. rhopaloccra male right antenna; 132-133, dorsal and lat- 
eral views of male A. mourei pronotum showing overlapping mesoscutum (A), acute lateral angle 
(B), and angular posterior lobe (C); 134, lateral view of female A. nasutus pronotum showing 
yellow maculation (stippling); 135, lateral view of female A. leunculus pronotum showing 
carina (arrow) not found on A. nasutus; 136-140, last visible (6th) male sternites: 136, A. 
intermedins; 137, A. boliviensis; 138, A. lanosus; 139, A. mourei; 140, A. inca. 



Revision of the Bee Genus Agapostemon 



581 




Figs. 141-154. Anterior views of male Agapostemon hind femur, tibia and basitarsus. Color 
yellow or creamy with brown or black maculations. Anterior maculations with left oblique or 
vertical hatching; posterior maculations with right oblique or horizontal hatching; anterior and 
posterior maculations with cross hatching. 141, A. \ohliellus; 142, A. atrocaeruleus ; 143, A. 
lennculus; 144, A. poeyi; 145, .-/. viequesensis; 146, A. jamaicensis ; 147, A. swainsonae; 148, 
A. hispaniolicus; 149, A. cubensis; 150, A. viridulus; 151, A. intermedins; 152, A. centratus; 
153, A. nasutus; 154, A. rhopdocera. Arrows indicate ventral views of basal ridge and apical 
groove on basitarsus. 



582 



The University Science Bulletin 




Figs. 155-166. Anterior views of male Agapostemon hind femur, tibia, basitarsus. Color 
yellow or creamy with brown or black maculations. Anterior maculations with vertical hatch- 
ing; posterior maculations with horizontal hatching; anterior and posterior maculations with cross 
hatching. 155, A. melliventris; 156, A. mexicanus; 157, A. peninsularis; 158, A. virescens; 159, 
A. tyleri; 160, A. coloradinus; 161, A. radialus; 162, A. cockerelli; 163, A. femoratus; 164, A. 
angelicus; 165, A. texanus; 166, A. splendens. Arrows indicate ventral views of basal ridge and 
apical groove on basitarsus. 



Revision of the Bee Genus Agapostemon 



583 




167-172 



173-177 



Figs. 167-172. Anterior views of male Agapostemon hind femur, tibia, basitarsus. Color 
yellow or creamy with brown or black maculations. Anterior maculations with left oblique 
hatching; posterior maculations with right oblique hatching; anterior and posterior maculations 
with cross hatching. 167, A. mowei; 168, A. inca; 169, A. lanosus; 170, A. boliviensis; 171, 
A. semimelleus; 172, A. chapadensis. Arrows indicate ventral views of basal ridge and apical 
groove on basitarsus. 

Figs. 173-177. Apical view of male gonostyli (right side): 173, A. chapadensis; 174, A, 
semimelleus; 175, A. inca; 176, A. mowei; 177A, A. boliviensis; 177B, A, lanosus, 



5S4 



The University Science Bulletin 




Figs. 178-196. Apical view of male gonostyli (right side): 178, Agapostemon viridulus; 
179, A. splendens; 180, A. texanus (A, apical stylus; B, medial plate; C, basal stylus); 181, 
./. angelicus; 182, A. centratus; 183, A. virescens; 184, A. tyleri; 185, A. coloradinus; 186, A. 
poeyi; 187, A. viequesensis ; 188, A. melliventris; 189, A. mexicanus; 190, A. peninsularis; 
191, A. \ohliellus; 192, A. intermedins; 193, A. leunculus; 194, A. jemoratus; 195, A. nasutus; 
196, A. atrocaeruleus. 



Revision of the Bee Genus Agapostemon 



585 




Figs. 197-214. Dorsal view of male genitalia (showing striae and pubescence on right side 
only): 197, Agapostemon virescens; 198, A. tyleri; 199, A. coloradinus; 200, A. splendent: 
201, A. texanus; 202, .-/. femoratus; 203, A. melliventris ; 204, A. mexicanus; 205, A. 
peninsidai is ; 206, A. centratus; 207, A. rhopedocera; 208, A. viridulus; 209, .-/. swainsonae; 
210, A. poeyi; 211, A. leunculus; 212, A. kphliellus; 213, A. atrocaeruleus ; 214, A. intermedins. 



586 



The University Science Bulletin 




215 




216 




217 




218 




219 




220 




215-223 



224-227 



Figs. 215-227. 215-221, dorsal view of male genitalia: 215, A. boliriensis; 216, A. lanosus; 
217, A. mourei; 218, A. inca; 219, A. chapadensis; 220, A. semimelleus; 221, A. nasutus; 222, 
lateral view of right penis valve of A. nasutus; 223, ventral view of male genitalia of A. 
nasutus (ventral flap removed on right); 224, posterior view of A. viequesensis male 
pygidium; 225, posterior view of A. ochromops male pygidium; 226, distal view of A. colo- 
radinus right penis valve; 227, distal view of A. tyleri right penis valve. 



Revision of the Bee Genus Agapostemon 



587 




Fig. 228. Distribution maps of the four North American species groups. The dendrograms 
indicate phenetic and presumed cladistic relationships within each species group. 



588 



The University Science Bulletin 



INDEX TO INCLUDED TAXA 

Valid Agapostemon species are in boldface. When different from current 
generic placement, the original genus of description is preceded by a semicolon. 
Page numbers are in boldface for descriptions or synonymies and in italics for 
illustrations. 



abjectus Cockercll, Agapostemon obscuratus, 
513. 

aeneus (Schrottky), Pseadagapostemon; Aga- 
postemon, 443. 

aenigma Roberts, Agapostemon, 445, 454, 458, 
460, 469, 476, 503, 514, 518, 559. 

aeruginosas Smith, Agapostemon, 529. 

alayoi Roberts, Agapostemon, 445, 457, 461, 
476, 514, 518,559. 

andensis (Vachal), Agapostemon: Halictus, 
525. 

Andrena, 441, 547, 550, 551. 

angelicus Cockerell, Agapostemon, 444-446, 
451, 453-455, 461-463, 503, 520, 522, 
534, 535, 537, 538, 555, 570, 571, 576, 

582, 584, 587. 

Apis, 441, 547, 550, 551, 554. 

arechavaletae (Schrottky), Pseadagapostemon; 

Agapostemon, 443. 
arenarius (Schrottky), Pseadagapostemon; 

Agapostemon, 442, 443. 
argentinus Holmberg, Agapostemon, 470. 
ascius Roberts, Agapostemon, 445, 448, 458, 

463-464. 
atcr Friese, Agapostemon nasutas, 495, 497. 
atrocaeruleus Friese, Agapostemon, 445, 448, 

449, 454, 455, 464-467, 481, 503, 528, 574, 

577, 581, 584, 585. 
Augochlora, 474. 
Aagochloropsis, 442. 
azarae Holmberg, Agapostemon, 470. 
bicolor Robertson, Agapostemon , 547. 
boliviensis Roberts, Agapostemon, 445, 449, 

459, 467-468, 483, 485, 494, 505, 580, 

583, 586. 

bonaerensis (Schrottky), Pseadagapostemon; 
Agapostemon, 443. 

borealis Crawford, Agapostemon, 533. 

brachycerus (Vachal), Agapostemon; Halictus, 
533. 

brachianas (Schrottky), Ruizantheda ?; Halic- 
tus, 443. 

bruneri (Crawford), Paragapostemon; Aga- 
postemon, 443. 

caelestina (Westwood), Paragapostemon; 
Nomia, 443. 

caeruleus (Ashmead), Aagochloropsis; Aga- 
postemon, 442. 

californicus Crawford, Agapostemon, 482, 533, 
537. 

castaneus Schrottky, Agapostemon, 470, 525. 

centratus (Vachal), Agapostemon; Halictus, 
445, 447, 453, 456, 460, 469-470, 476, 478, 
514, 518, 544, 559, 579, 581, 584, 585. 



chalets (Vachal), Agapostemon; Halictus, 517. 

chapadensis Cockerell, Agapostemon, 445, 448, 
449, 458, 459, 470-471, 526, 583, 586. 

chiriqaiensis (Vachal), Agapostemon; Halictas, 
554. 

cillaba (Cameron), Paragapostemon; Nomia, 
443, 471. 

citricornis (Vachal), Pseadagapostemon ; Halic- 
tus, 442. 

Clementinas Cockerell, Agapostemon californi- 
cus, 533, 534, 537. 

cockerelli Crawford, Agapostemon, 445, 451, 
453-455, 471-473, 482, 503, 520, 542, 560, 
575, 578, 582, 587. 

coloradensis Crawford Agapostemon 473, 542. 

coloradensis Titus, Augochlora, 474. 

coloradinus (Vachal), Agapostemon; Halictus, 
445, 446, 452, 473-476, 500, 542, 543, 548, 

561, 575, 578, 582, 584-587. 

columbi Roberts, Agapostemon, 445, 457, 458, 
460, 476-477, 478, 486, 514, 518, 544, 559. 

coryliventris Holmberg, Agapostemon, 525. 

Corynura, 443. 

cubensis Roberts, Agapostemon, 445, 456, 476, 
478-479, 514, 544, 559, 579, 581. 

cyaneus Roberts, Agapostemon, 445, 457, 476, 
478, 479-480, 514, 518, 524, 525, 544, 559. 

cyanozonus Cockerell, Agapostemon, 533. 

digaeti Cockerell, Agapostemon, 499. 

dimidiatus (Lepeletier), Agapostemon; Halic- 
tus, 547. 

divaricatus (Vachal), Pseadagapostemon; 
Halictus. 443. 

emarginatus (Spinola), Ruizantheda; Halictus, 
443. 

erebus Roberts, Agapostemon, 445, 448, 455, 
463, 480-481, 503. 

experiendus Holmberg, Agapostemon, 470. 

fasciatus Crawford, Agapostemon, 499. 

fasciatus Nylander, Halictus, 499. 

femoralis (Guerin), Agapostemon; Andrena, 
441, 550, 551. 

femoratus Crawford, Agapostemon, 442, 445, 
451, 453, 471-473, 482-483, 503, 520, 522, 

562, 575, 578, 584, 585, 587. 
festivus Cresson, Agapostemon , 517. 
gualanicus Cockerell, Agapostemon nasutus, 

508. 
Halutus. 441-444, 473, 474, 499, 517, 525, 

529, 533, 547, 554. 
heterurus Cockerell, Agapostemon, 445, 448, 

449, 458, 468, 483-484, 486, 494, 505. 
hispaniolicus Roberts, Agapostemon, 445, 456, 

457, 460, 461, 476-478. 484, 514, 518, 519, 

532, 544, 546, 551, 559, 574, 579, 581. 



Revision of the Bee Genus Agapostemon 



589 



idahoensis Michener, Agapostemon angelicas, 

533, 534. 
inca Roberts, Agapostemon, 445, 449, 459, 

468, 484-485, 505, 580, 583, 586. 
insularis Roberts, Agapostemon, 445, 456, 458, 

460, 476, 478, 485-486, 491, 514, 518, 544, 

559. 
intermedius Roberts, Agapostemon, 445, 448, 

449, 453, 455, 458, 459, 464, 465, 467, 468, 

485, 486-490, 503, 505, 575, 577, 580, 581, 

584, 585. 
iowensis Cockerell, Agapostemon texanus, 533. 
jamaicensis Roberts, Agapostemon, 445, 456, 

476, 478, 485, 490-491, 514, 518, 544, 559, 

581. 
joseantis Friese, Agapostemon, 533. 
kohliellus (Vachal), Agapostemon; Halicttis, 

445, 447, 456, 457, 461, 476, 478, 491-494, 

514, 544, 559, 574, 579, 581, 584, 585. 
lanosus Roberts, Agapostemon, 445, 449, 459, 

468, 484, 485, 494-495, 505, 580, 583, 586. 
Lasioglossum, 444. 
leunculus Vachal, Agapostemon, 442, 445, 

448, 451, 453-455, 458, 459, 463, 464, 481, 

487, 495-499, 500, 503, 515, 523, 524, 554, 

563, 574, 577, 580, 581, 584, 585. 
martini Cockerell, Agapostemon, 471, 474, 

542. 
melanurus Cockerell, Agapostemon, 508. 
melliventris Cresson, Agapostemon, 445, 452, 

454, 455, 469, 487, 496, 499-503, 504, 515- 

517, 520, 564, 575, 577, 582, 584, 585, 587. 
mexicanus Roberts, Agapostemon, 445, 447, 

452-455, 496, 503-504, 515, 576, 577, 582, 

584, 585, 587. 
mourei Roberts, Agapostemon, 445, 449, 459, 

467, 468, 483-485, 494, 495, 504-508, 580, 

583, 586. 
multicolor Holmberg, Agapostemon, 525. 
mtttabilis (Spinola), Ruizantheda (Ruizan- 

thedella); Halicttis, 443. 
nasita Schrottky, Pseudagapostemon , 442. 
nasutus Smith, Agapostemon, 439, 442, 445, 

448, 451-453, 455, 458, 459, 463, 464, 481, 

492, 496-500, 503, 508-513, 515, 524, 554, 

565, 574, 577, 580, 581, 584, 586. 
nigricomis (Fabricius), Agapostemon; An- 

drena, 547. 
Nomia, 441, 443, 471, 517. 
obscuratus Cresson, Agapostemon, 445, 456, 

457, 476, 478, 484, 513, 514, 518, 544, 551, 

559. 
ochromops Roberts, Agapostemon, 445, 457, 

460, 476, 478-480, 513-515, 518, 524, 525, 

544, 559, 586. 
olivaceo-splendtns (Strand), Pseudagaposte- 
mon; Agapostemon, 443. 
Paragapostemon, 441-443. 
paidista Schrottky, Pseudagapostemon, 442. 
peninsularis Roberts, Agapostemon, 442, 445, 

452-455, 496, 500, 503, 504, 515-517, 576, 

577, 582, 584, 585, 587. 
peruvianus Cameron, Agapostemon, 508. 



pissisi (Vachal), Pseudagapostemon; Halicttis, 
443. 

placidus (Smith), Ruizantheda; Halicttis, 443. 

plunfasciatus (Vachal), Agapostemon; Halic- 
ttis, 499. 

podager (Vachal), Paragapostemon; Halicttis, 
442. 

poeyi (Lucas), Agapostemon; Andrena, 445, 

447, 456, 458, 460, 476-478, 485, 486, 491, 
514, 515, 517-519, 525, 544, 546, 559, 566, 
567, 581, 584, 585. 

portoricensis Cockerell, Agapostemon radi- 

atus, 544. 
proscriptelltts Cockerell, Agapostemon, 533, 

534. 
proscriptus Cockerell, Agapostemon, 533. 
proximus (Spinola) Ruizantheda; Halicttis, 

443. 
psammobius Cockerell, Agapostemon calijorni- 

cus, 533, 534. 
Pseudagapostemon , 442, 443. 
pulchra Smith, Agapostemon, 520. 
purpweopictus Cockerell, Agapostemon , 508. 
radiatus (Say), Agapostemon; Halicttis, 439, 

445, 451, 453, 472, 473, 482, 483, 520-523, 

547, 568, 575, 578, 582, 587. 
rhopalocera Smith, Agapostemon, 445, 447, 

448, 453, 496, 523-524, 576, 580, 581, 585. 
rohinsoni Cresson, Nomia, 517. 
Ruizantheda, 443. 

(Ruizanthedella) , Ruizantheda, 443. 
sapphirinus Roberts, Agapostemon, 445, 457, 
476, 478, 479, 514, 518, 524-525, 544, 559. 
semimelleus Cockerell, Agapostemon, 445, 448, 

449, 458, 459, 470, 471, 486, 525-528, 579, 
580, 583, 586. 

semiriridis Cresson, Agapostemon , 551. 

sericea Forster, Apis, 547, 554. 

sicheli (Vachal), Paragapostemon; Halicttis, 
443. 

splendens (Lepeletier), Agapostemon; Halic- 
ttis, 439, 444-446, 451, 452, 454, 476, 503, 
520, 529-531, 535, 569, 576, 582, 584, 585, 
587. 

subtilior Cockerell, Agapostemon texanus, 533, 
534. 

sulcatulus Cockerell, Agapostemon, 520. 

suljuripes Friese, Agapostemon, 533, 534. 

swainsonae Cockerell, Agapostemon, 445, 456, 
457, 476, 478, 484, 514, 518, 531-533, 544, 
551,559, 574, 579, 581, 585. 

tacita (Cameron), Paragapostemon; Nomia, 
443. 

texanus Cresson, Agapostemon, 439, 444-446, 

451, 453-455, 461-463, 465, 466, 472, 473, 
475, 482, 483, 488, 489, 497, 501-503, 515, 
520-524, 529, 530, 533-542, 543, 548, 550, 
555, 570, 571, 576, 582, 584, 585, 587. 

tricolor (Lepeletier), Agapostemon; Halictus, 

547. 
tyleri Cockerell, Agapostemon. 442, 445, 447, 

452, 454, 455, 474, 503, 515, 542-544, 548, 
572, 573, 575, 578, 582, 584-587. 



590 



The University Science Bulletin 



vandykpi Cockerell, Agapostemon texamis, 
533-535, 537. 

viequesensis Cockerell, Agapostemon, 445, 447, 
457, 460, 476-478, 480, 486, 491, 514, 515, 
518, 519, 525, 544-547, 559, 573, 574, 579, 
581, 584, 586. 

virescens (Fabricius), Agapostemon; Andrena, 
445, 451, 452, 474-476, 542, 543, 547-550, 
551, 554, 575, 578, 582, 584, 585, 587. 



viridulus (Fabricius), Agapostemon; Apis, 
441, 445, 447, 456, 457, 461, 469, 476, 478, 
479, 484, 513, 514, 518, 532, 544, 550-554, 
559, 574, 579, 581, 584, 585. 

I'ltlpicolor Crawford, Agapostemon, 495. 

xanthorhinus (Cockerell), Psendagapostemon; 
Agapostemon, 442. 

zosteronedys (Moure), Psendagapostemon; 
Agapostemon, 443. 



THE UNIVERSITY OF KANSAS 

SCIENCE BULLETIN 

Vol. XLIX Pages 591-729 March 10, 1972 No. 10 



TABLE OF CONTENTS 
ABSTRACT 593 

INTRODUCTION 593 

ACKNOWLEDGMENTS 597 

MATERIALS AND METHODS 598 

ACCOUNT OF SPECIES 600 

Family Emballonuridae 602 

Family Noctilionidae 608 

Family Phyllostomatidae 615 

Chilonycterinae 615 

Phyllostomatinae 622 

Glossophaginae 629 

Carollinae 636 

Sturnirinae 643 

Stenoderminae 649 

Desmodontinae 657 

Family Natalidae 664 

Family Vespertilionidae 669 

Family Molossidae 682 

HISTOCHEMISTRY OF THE GASTRIC MUCOSA AND 
DUODENAL JUNCTION 690 

GASTRIC MORPHOLOGY IN RELATION TO FEEDING . .... 702 

SYSTEMATIC RELATIONSHIPS AS REVEALED BY 
GASTROINTESTINAL STUDIES 

LITERATURE CITED 727 



Comparative Morphological and Histochemical 
Studies of Stomachs of Selected American Bats 

G. Lawrence Forman* 

Museum of Natural History 
The University of Kansas 

ABSTRACT 

Stomachs of 13 species of North American hats representing genera of six 
families and several different types of feeding habits were examined and com- 
pared morphologically, histologically, and histochemically. All species studied 
were found to have morphologically distinct stomachs; relationships hetween 
gastric structure and physiology, and between gastric structure and consumption 
of particulate food materials were investigated. Four histochemical procedures 
were employed to examine qualitative and quantitative features of mucins in the 
gastric mucosa and the glands of Brunner. Stomachs of insectivorous and carniv- 
orous bats studied generally were of simpler structure than the more specialized 
stomachs of frugivorous and sangivorous kinds, with various degrees of inter- 
mediacy in structure noted in two more or less omnivorous species. Results of 
histochemical studies indicate greatest overall production of mucus in insecti- 
vorous and carnivorous bats. 

Some systematic relationships among families, subfamilies, and genera are 
discussed in light of results of gastric studies. Apparent coevolution of gastric- 
morphology with a progressive change from albuminous to cellular feeding 
within the family Phyllostomatidae also is discussed. 

INTRODUCTION 

Definitive phylogenetic studies of bats have been hampered by lack of 
an adequate fossil record for this group. The earlier-supposed relationships 
(Dobson, 1875; Miller, 1907) of major taxa in the Chiroptera have been 
questioned in recent years, leading to a search for, and subsequent use of, 
a whole suite of approaches and characteristics not heretofor employed 
in the classification of these mammals. The purpose of this study was to 
examine the structure of the gastric portion of the intestinal tract for possible 
use as a criterion in studying relationships among bats. Additionally, be- 
cause mammalogists and mammalian paleontologists have long used selected 
structures of the teeth, palate, and masticatory musculature in examining 
coevolution of morphology and food habits in various groups of mammals, 
this study also attempts to correlate gastric morphology with food habits. 

The objectives of the present study were fourfold. The first aim was to 
increase our knowledge of the gastric structure of bats along with detailed 
comparisons of the gastric mucosae. Second, some histochemical properties 

♦Present address: Department of Biology, Rockford College, Rocktord, Illinois 61101. 



594 The University Science Bulletin 

of the gastric and duodenal mucins were examined and described. Third, the 
results of the above investigations were evaluated with respect to diverse and 
usually obligate feeding habits of various bat species, in an attempt to 
demonstrate the presence or absence of coevolution in feeding preferences 
and gastric structure. Lastly, gastric morphology and distribution of various 
mucosubstances (as revealed by selected histochemical procedures) were 
considered in the light of the presently recognized phylogenetic arrange- 
ment of the species under examination. 

Published comparative studies of mammalian soft anatomy are, for the 
most part, inadequate, because the variability among species is rather poorly 
documented. Deficiencies are especially apparent for the post-lingual por- 
tions of the gastrointestinal tract of mammals. With the exception of a 
few accounts dealing with domesticated animals and various groups of 
rodents and insectivores, most comparative anatomical studies of stomachs 
make reference to only a few species, or concern specific physiological prob- 
lems in specific species. Such studies have left gaps in our knowledge of 
how much stomachs of relatively closely related taxa differ structurally and 
physiologically. Also, because there are few comparative studies of the 
gastric mucosal membrane, relatively little has been suggested regarding the 
relationship of gastric structure to feeding habits, other than a few well- 
entrenched generalizations about certain domesticated mammals. This paper 
presents new information regarding the coevolution of food habits with 
gastric structure in bats. 

According to Cuvier (1805), there were three distinguishable forms that 
could be recognized; the transversely elongated tubular stomach possessed 
by bats that eat fruit, the globular stomach with closely approximated cardiac 
and pyloric orifices exhibited by insectivorous bats, and the elongated 
"conical" stomach, with a specialized pyloric caecum, of blood feeders. The 
last of these categories is immediately questionable because of the fact that 
the species upon which the category is based were probably not blood- 
feeding species, but nectar- or fruit-eaters. Huxley (1865), subsequent to 
reviewing Cuvier's classification, described the stomach of the truly blood- 
feeding vampire bat (Desmodns), in which he found the cardiac portion 
to be a tremendously elongated caecum, the cardiac and pyloric orifices 
close together, and the region of the pyloric mucosa notably abbreviated. 
Owen (1868) gave his own interpretation of various types of stomachs of 
bats in relation to diet, including one poorly denned category concerning 
"times of taking the food and to the quantity taken." 

The important work of Robin (1881) represents the first detailed pub- 
lished comparative account of gastrointestinal structure in bats. Robin's 
paper includes an excellent review of earlier work, along with new infor- 
mation that led him to discard the classifications of Cuvier and Owen as 



Studies of Stomachs of Selected American Bats 595 

inconsistent with his own observations. Robin noted three shapes of stom- 
achs as demonstrated by the 1) Megachiroptera (Pteropodidae), 2) Micro- 
chiroptera (Nyctimene) and 3) Desmodus (distinctive type of vampire 
bats). He also discussed variation between species in orientation of mucosal 
folds and variation in distribution of certain cell types. He also noted the 
lack of a "portio cardiaque" in the insect-eating Nyctimene cephalotes, a 
structure characteristically present in fruit-eating megachiropterans. Grasse 
(1955) reiterated the basic classification of Robin, and additionally noted 
differences between the vampire genera Desmodus and Diphylla in the 
lengths of the caecal pouches, and that the distribution and structure of 
gastric glands were notably different in Desmodus from those of other bats. 

Several early authors, notably Rollett (1871) and Langley (1880-82), de- 
scribed the structure of the gastric mucosa of bats, but, unfortunately, ne- 
glected to name the kinds examined. Langley noted that the fundic glands 
were generally short, that those of the greater curvature were especially 
long, that parietal cells were relatively few in number, and, although perhaps 
unwittingly, was probably the first to describe a pylo-fundic transition zone. 
His observations seem to agree well with those included here for Plecotus 
townsendii. Rollett's study was similar to Langley's, although Rollett 
claimed to have observed reduced numbers of "adelomorphous" or border 
cells in summer-taken as opposed to winter-taken specimens. Present evi- 
dence questions the reliability of these findings. 

Until relatively recently, the bulk of investigation concerning gastroin- 
testinal structure in the Chiroptera was almost entirely macroscopic in 
nature. Several noteworthy earlier papers deserve consideration. Mathis 
(1928a), in an extensive treatise restricted to a discussion of the intestinal 
portion of the digestive tract, described the patterns of intestinal folds and 
selected glandular areas in the intestines of nine species of bats, and noted 
differences in intestinal length and relationships between patterns of in- 
testinal folds and types of nutrition. Eisentraut (1950) noted that among 
the species he examined the stomachs of insect-eaters tended to be simpler 
and less saccular than those of the megachiropteran fruit-eaters, and that 
the tendency toward the development of an extensive blind-ended, saccular 
fundic caecum could be found only in the Phyllostomatidae of the Micro- 
chiroptera. He also noted that two glossophagine bats, Glossophaga and 
Choeronycteris, had stomachs that closely resembled the condition in insect- 
eating vespertilionid and emballonurid species. However, Eisentraut's illus- 
trations of stomachs of some species strongly suggest that they were dis- 
tended and therefore not typical of the "empty" condition; his drawings of 
stomachs of Carollia and Artibeus are, however, nearly identical to those 
included for these genera herein. Fischer (1909) established that the mucous 
membrane was better developed in the cranial than the caudal portion of 



596 The University Science Bulletin 

the stomach. The same was stated for the musculature, but, unfortunately, 
no reference was made as to the species examined. Fischer held that a 
condition of localized hypertrophied mucous membrane and musculature 
was traceable to an adaptation to eating hard insects. Kolb (1954) examined 
stomachs of Rhinolophus, Nyctalus, and Old World Eptesicus, and consid- 
ered reductions and specializations of certain portions of the gastric mucosa 
in these bats in relation to diets predominantly composed of insects. Kolb 
concluded that specializations such as folding of the stratified epithelium 
at the gastroesophageal junction, greater development of the musculature 
and gastric mucosa in the lesser as opposed to the greater curvature (as 
noted by Fischer, 1909), and secondary reduction of gastric mucosa in the 
blind end of the caecum were in some way related to insectivorous feeding 
habits, or to protection of the mucosal membrane from abrasion by chi- 
tinous fragments. 

Park and Hall (1951) examined the tongues and stomachs of eight 
species of New World bats, commenting on the apparent specializations in 
both tongue and stomach to incorporate a heavily liquid diet in Desmodus. 

Studies that deal specifically with the histological structure of the gastro- 
intestinal tract of bats are few, and even fewer are comparative in approach. 
Fischer (1909) examined microscopic sections of the gastrointestinal wall, 
showing for Barbastella and Vespertilio that the musculature was thicker 
within the lesser curvature. Fischer was the first worker to examine in detail 
the gastroesophageal junction (Barbastella). He noted that the junction 
between the esophageal epithelium and gastric mucosa was abrupt and 
described a local thickening of the circular muscle layer at the junction 
similar to that described herein for Rhynchonycteris naso. Mathis (1928b) 
examined the cellular structure of the glands of Brunner of three vesper- 
tilionid species and a rhinolophid. He observed in all cases a clear-cut dis- 
tinction in cellular morphology between the glands of Brunner and the 
pyloric glands. Mathis noted slight differences in Brunner's glands in dif- 
ferent species; his figure of the pyloric junction in Vesperugo (=Pipistrel- 
lus) pipistrellus is notably similar to that of Plecotus included here. Kolb 
(1954) examined the cardiac glands of two bats (probably Nyctalus and 
Rhinolophus), concluding that they probably functioned as a source of lubri- 
cant tor materials entering the stomach, and also observed in Rhinolophus 
hipposideros that the cardiac glands penetrated the lower esophageal epi- 
thelium within a narrow belt at the gastroesophageal junction. An especially 
important recent contribution is that of Schultz, 1970. This paper proposes 
familial status for a megachiropteran, Harpionycteris whiteheadi, based on 
studies of gastrointestinal structure. 

A recent paper by Rouk and Glass (1970) presents comparative accounts 
<>l gastric structure in five species of North American bats, and reviews 
additional pertinent literature not referred to here. 



Studies of Stomachs of Selected American Bats 597 

Two significant contributions concerning comparative gastrointestinal 
morphology, although not specifically considering chiropteran anatomy, arc 
those of Allison (1948) and Myrcha (1967) on the Insectivora. These papers 
include comprehensive accounts of gross gastric morphology of both Old 
and New World insectivores, and I have referred to them occasionally for 
comparative purposes, because the Insectivora and Chiroptera are basically 
similar in several aspects of gross gastric structure. 

Histochemical examination of intestinal sections was undertaken in an 
attempt to examine the qualitative and quantitative nature of mucopolysac- 
charides in the gastric mucosa and the submucosal glands of Brunner. 
Owing to the limited amount of material available for these studies, and 
to the nature of the fixatives employed, only a few techniques were used 
and the study must be regarded as preliminary in nature. Longitudinal 
sections of stomachs examined histologically were included in all mucin 
reactions when possible. 

Prior to the relatively recent development of procedures for differen- 
tially staining and identifying mucous substances, it long had been recog- 
nized that mucins of various cells of the gastrointestinal epithelium reacted 
differently, and that differences in staining properties of the same cell type 
could be found in different species. Clara (1940) noted staining differences 
between types of mucous cells specifically with reference to stomachs of 
different species, and Liebermann (1SS7-SS, as cited by Jennings and Florey, 
1956) had previously suggested, on the basis of chemical investigations on 
extracts of mucosa, that different kinds of mucin were present in mammalian 
stomachs. Liebermann categorized mucins into those that contained sulphur, 
and those that did not. It is of some interest to note that present-day classi- 
fications of mucopolysaccharides (see Spicer, 1963) employ the condition 
of the presence or absence of sulfate radicals. A few workers, including 
Clara (1940) and Gomori (1952), have suggested that gastrointestinal mucins 
may be distinguished as either true mucin or mucoid on both histological 
and histochemical grounds, although different techniques of fixation and 
staining apparently yield inconclusive results (Jarvi and Lauren, 1951). A 
comprehensive review and summary of the best known and most widely 
used procedures for characterizing mammalian mucopolysaccharides is that 
of Spicer (1963). 

ACKNOWLEDGMENTS 

I am especially indebted to Dr. J. Knox Jones, Jr., for his constant assistance 
during the course of this study, and to Dr. William Coil for helpful suggestions 
regarding technique and histochemical procedure. I am grateful also to Drs. 
Jones, Coil, and William M. Balfour for critical review of the manuscript; to Dr. 
William E. Duellman for the loan of microscopic equipment; to Ronald W. 
Turner, Timothy E. Lawlor, Hugh H. Genoways, Duane A. Schlittler, and Dr. 



598 The University Science Bulletin 

Roderick Suthers for providing gastrointestinal tracts of bats; and to Carleton J. 
Phillips and James D. Smith for providing stimulating discussion and criticism 
during preparation of the manuscript. 

Support for this study was provided through an award to me from the 
Department of Zoology at The University of Kansas, Lawrence, and by way of a 
grant (no. 4193-5706-2) to Dr. Jones from the Bio-Medical Research Fund at 
the university. Many Neotropical specimens were collected under the aegis of a 
contract (DA-49-193-MD-2215) from the U.S. Army Medical Research and De- 
velopment Command. 

MATERIALS AND METHODS 

Whole gastrointestinal tracts of bats were taken from specimens collected in nets or shot 
during crepuscular feeding, and, in one case (Plecotus townsendii) , from torpid individuals 
taken in a winter hibernaculum. Most tropical species were netted, and usually were fasted 
for a minimum of 10 hours. A total of 28 stomachs, chosen because they were adequately 
preserved and devoid of gastric contents, were used for histological and histochemical exami- 
nation. An additional 12 specimens, including one of Noctilio leporinus (a species of which I 
had no suitable specimens) were examined only macroscopically. Specimens listed as examined 
were those sectioned and stained. Descriptions of topography include observations made on 
non-sectioned material, whereas the recorded measurements are only from stomachs prepared 
for histological examination. 

The exact physiological state of specimens at the time of preservation is unknown, al- 
though, if used for histological examination, specimens had stomachs devoid of gastric contents. 

All material was fixed in either Bouin's or 8% formalin in the ratio of at least 10 parts 
fixative to one part tissue by volume, and subsequently preserved in 70% ethyl alcohol. 
Several stomachs from specimens preserved in spirits as museum specimens were sectioned 
and stained with hematoxylin and eosin, but judged unsuitable for microscopic study. 

Material was imbedded in 52.0° Tissuemat and sectioned frontally at 5-7 microns. The 
following histological techniques were employed: 

1) Harris' hematoxylin and eosin (Harris' formula after Davenport, 1960). 

2) Mallory's triple connective tissue stain (Pantin, 1948). 
The following histochemical procedures were used: 

1) Toluidine blue O (Lillie, 1965).- — A 1:1000 aqueous solution was employed with ex- 
posure for 15 seconds with immediate transfer to acetone. 

2) Steedman's Alcian blue 8GX for mucins (Steedman, in Lillie, 1965). — Alcian blue 
8GX 300 (Allied Chemical Co.) was employed for 60 seconds with a distilled water rinse. 
This procedure employs an aqueous solution of the dye and was used at a pH of 7.1. 

3) Maxwell's periodic acid Alcian blue for gastric mucus (Maxwell, in Lillie, 1965). This 
procedure, which employs exposure of sections to periodic acid oxidation prior to staining was 
used as a supplement to Steedman's method, and as an additional check for glycogen. 

4) Periodic Acid SchifT sulfite leucofuchsin reaction (Lillie, 1965). Control slides (no ex- 
posure to periodic acid) also were employed. 

In addition to successful histochemical procedures described above, several other types were 
attempted, but did not provide conclusive results. One of these was digestion with testicular 
hyaluronidase, followed by toluidine blue O (Lillie, 1065). Sections exposed to 18 hours 
digestion in 1:10,000 aqueous hyaluronidase at 37°C revealed no distinguishable alteration 
ot toluidine blue metachromasia in comparison to control sections, therefore yielding negative 
results in a test for chondromucins and hyaluronic acid. 

The two Alcian blue procedures were run as single series, one with, and one without, 
periodic acid oxidation prior to staining. The toluidine blue O procedure was carried out 
twice, once as a single series, then as a control series for the hyaluronidase lysis. 

Scoring of the intensity of Alcian blue and PAS reactions was as follows: exceptionally 
strong (-}--)- _)--(-), moderately strong ( + + + ), moderate ( + + ), weak (+), barely precepti- 
ble (trace), not observed ( — ). Toluidine blue O was scored simply as plus ( + ), the presence 
of demonstrable metachromasia, and minus ( — ), metachromasia not observed. 



Studies of Stomachs of Selected American Bats 



599 



VENTRAL 
VIEW 



pyloric constriction 



sulcus intermedius 



RIGHT 



sulcus angulans 




cardiac vestibule 



LEFT 



fundic caecum 



POSTERIOR 



CRANIAL 
VI EW 

DORSAL 



RIGHT 




LEFT 



VENTRAL 

Fig. 1 (upper). Diagrammatic representation of the stomach of Carollia perspicillata (ven- 
tral view). Note the location of various sulci. 

Fig. 2. (lower). Stomach of C. perspicillata (diagrammatic), viewed cranially. 



Anatomical terminology and descriptions of location and direction of parts used with 
reference to the gastrointestinal tract frequently are undefined or unclear, sometimes inap- 
propriate, and otherwise variously misused in the literature. For these reasons, a list of defined 
terms is included here in addition to the brief discussion of orientational terminology to follow. 

All descriptions and discussions of gastric morphology are based on location of the stomach 
in situ in the abdominal cavity. Anterior (craniad) and posterior (caudad) are here con- 
sidered those directions of the lesser and greater curvatures, respectively (Fig. 1), usually con- 



600 The University Science Bulletin 

sidered dorsal and ventral, respectively, in the literature. Dorsal is here considered that surface 
of the stomach nearest the vertebral column, whereas ventral is that surface farthest from (or 
opposite to) the vertebral column (Fig. 2). All gastric structures are distal to the apex of the 
fundic caecum, and likewise proximal to the pyloric sphincter. 

All measurements of gross structures are in millimeters, and are included to give an 
indication of stomach size and orientation of various features with respect to others. Measure- 
ments of depths of the various zones of gastric mucosa were obtained using a dial micrometer. 
The ranges are recorded in microns. 

All photographs are of structures stained with Harris' haematoxylin and eosin unless other- 
wise indicated. 

Mid-longitudinal outline drawings represent composites of specimens of each species, and 
were produced by projecting representative sections onto paper using a photographic enlarger, 
tracing the outlines for each specimen, and producing a composite outline by means of a light 
table. Within some mid-longitudinal drawings, structures not observed in medial sections, 
although important enough to be included in the illustrations, are depicted by broken lines. 

Uncommon morphological terms used in the accounts that follow are listed and defined 
below. 

Cardia. — Limited area of stomach surrounding gastroesophageal junction. 

Cardiac vestibule (same as cardia). — Part of stomach between gastroesophageal junction and 

general plane of lesser curvature. 
Elastic stomach. — That region of the stomach lateral to gastroesophageal junction terminating 

in a blind end, usually containing some elastic fibers. 
Fundic caecum. — Generally, area between gastroesophageal junction and left lateral terminus 

of stomach. 
Fundus. — That part of the stomach that secretes acid and pepsin, lying between cardiac and 

pyloric parts. 
Incisura cardiaca. — Notch formed between cardiac vestibule and fundic caecum at entrance of 

esophagus into stomach. 
Pyloric constriction. — External furrow at pyloric sphincter. 
Pyloric bend. — Cranial recurvature or arching of terminal stomach. 
Pylorus. — Glandular region of stomach composed entirely of cells resembling mucoid neck 

cells, located proximal to pyloric sphincter. 
Saccular stomach. — That portion between gastroesophageal junction and apex of fundic 

caecum. 
Sulcus angularis. — Furrow (external) formed generally at, or slightly proximal to, pyloric bend. 
Sulcus intermedius. — Furrow (external) approximately midway between pyloric bend and 

pyloric sphincter. 
Terminal stomach. — That portion between gastroesophageal junction and pyloric sphincter. 
Transition zone. — Zone of transitional gastric glands between fundic mucosa and other glandu- 
lar types. 

ACCOUNT OF SPECIES 

The gastrointestinal tracts of all Chiroptera examined consisted of the 
same typical segments that have been described for other kinds of mammals: 
a thin-walled, extremely narrow esophagus lacking submucosal or mucosal 
glands and lined entirely with stratified squamous epithelium, within which 
the usual four strata of squamous cells could be observed; a distinctive, 
saclike, muscular stomach (with the exception of the blind tube of Des- 
modus rotundus) completely lined with well-differentiated, usually simple, 
branched, tubular mucosal glands which are differentiated by the occurrence 
of particular cell types and identifiable within specific zones; a relatively 
long small intestine (especially long in Desmodits) with thin, muscular 



Studies of Stomachs of Selected American Bats 601 

walls, little loose submucosal tissue, and long, narrow villi; and a well- 
developed colon including numerous goblet cells and somewhat more ex- 
tensive non-muscular submucosa than the small intestine. 

The wall of the chiropteran stomach is composed of the tunica mucosa, 
tunica submucosa, tunica muscularis, and tunica serosa. Tunica serosa lines 
the entire external surface of the stomach; no squamous tissue is present 
except in the uppermost portion of the cardiac vestibule of Natalus stra- 
mineus. The amount of non-muscular submucosa in the stomachs of most 
bats examined was found to be extremely small in comparison to many 
other mammals. Infrequent to numerous accumulations of lymphatic tissue 
within the tunica propria were observed in some, but not all, specimens. 
Vascularization of the tunica propria was similar in all species examined 
with the exception of Desmodus rotundas, in which the complements of 
arterioles and venules are extensive. 

In all species the tunica mucosa, submucosa, and thin sheet of smooth 
muscularis mucosa just beneath the gastric glands are thrown into longi- 
tudinal or transverse (or both) and anastomosing rugae. These rugae, few 
to many depending on the species, extend throughout the stomach, termi- 
nating at the gastroesophageal and pylo-duodenal junctions. Rugae may or 
may not be reduced at the apex of the elastic stomach (fundic caecum), 
and are often poorly defined in a fully distended stomach. The tunica 
muscularis proper consists of an inner circular and outer longitudinal layer, 
with extensive variation in depths of layers in various species. A third, 
extremely thin layer of smooth muscle fibers lies to the inside of, and 
oblique to, the two outer layers, usually separated from them by a thin 
zone of loose submucosal tissue. 

In all species examined, I noted a partial or pronounced cranial recurva- 
ture of the aboral end of the stomach, often beginning substantially proximal 
to the pyloric sphincter (Natalus perhaps an exception). This condition has 
been observed in other species of bats, including members of the Mega- 
chiroptera, and immediately distinguishes stomachs of Chiroptera from 
many members of the Insectivora, notably soricids examined by Myrcha 
(1967). Curvature of the stomach is known in some Insectivora (Erina- 
ceidae, Talpidae) and its significance as a diagnostic character in the 
Chiroptera will be discussed later. 

All species examined (except Desmodus) have four distinctively recog- 
nizable zones of gastric mucosa — cardiac, fundic, pylo-fundic transition, and 
pyloric. Although Desmodus may in fact have all four zones represented, 
mucous glands occupying the area between oral and aboral orifices of the 
stomach are identical and are here considered to be pyloric glands. Cardiac 
glands are mucus-producing and appear to contain a single cell type. Fundic 
mucosa usually occupies the majority of the inner surface area and is com- 



602 The University Science Bulletin 

posted of tubular glands containing mostly zymogenic cells that are the 
proposed sites of pepsin secretion, parietal cells that secrete hydrochloric 
acid, and various mucus-producing cells. The transitional zone is a highly 
variable area, both dimensionally and structurally, representing the change 
from fundic to pyloric mucosa within the terminal stomach. The pyloric 
zone, always proximal to the pyloric sphincter, is a second mucus-producing 
area composed of relatively straight tubular glands with a single cell type. 

The stomach of the vespertilionid Plecotus townsendii was considered to 
be relatively generalized among bats examined. Because of its notable sym- 
metry and seeming lack of unusual specialization, the stomach is often 
used in the accounts in comparison or contrast to structures in other species 
where clarity by means of comparison appeared necessary. Species are 
arranged in the following accounts in phylogenetic order following Hall and 
Kelson (1959). 

Due to the small sample sizes of species employed in this study, the 
conclusions drawn here must be regarded as preliminary, and the study of 
a pilot nature. Further study utilizing many more specimens is encouraged, 
in order to document intraspecific variation. 

Family EMBALLONURIDAE 

Rhynchonycteris naso (Wied-Neuwied, 1820) 

Specimens examined (2). — Two specimens from Nicaragua (KU 105911, 
105915). 

Overall dimensions. — Greatest length 7.4-8.1; greatest breadth 3.1-3.8; 
gastroesophageal junction to pyloric sphincter 2.4-2.9; gastroesophageal junc- 
tion to apex of fundic caecum 2.5-2.8; breadth pylorus at sphincter 1.3-1.4. 

General morphology. — Stomach tubular, reniform, generally symmetrical 
although tapering gradually throughout from fundic caecum to pylorus; 
pronounced curvature in frontal plane throughout body with apex of fundic 
caecum directed antero-laterad; tubular stomach gradually curved anteriad 
two-thirds distance from apex of fundic caecum to pyloric sphincter, con- 
tinuing craniad as narrow, relatively long, tubular endpiece directed cranially 
and parallel to cardiac vestibule; lesser and greater curvatures nearly parallel 
throughout; terminal endpiece lateral to pyloric curvature largest among 
insectivorous species examined; stomach rounded on sides, convex on both 
dorsal and ventral surfaces. Cardiac vestibule present although short and 
narrow, musculature of wall well developed as through much of lesser 
curvature; vestibule at marked angle to lesser curvature and relatively 
uniform in breadth with little expansion posteriorly; incisura cardiaca rela- 
tively deep and narrow, created by marked angle of cardiac vestibule to 
lesser curvature; distinctive local thickening of both longitudinal and circu- 
lar muscle layers within incisura, but no apparent separation of circular 



Studies of Stomachs of Selected American Bats 



603 




2mm 



Fig. 3. Mid-longitudinal representation of the stomach of Rhynchonycteris naso. Explanation 
of symbols: CG, zone of cardiac glands; BG, zone of Brunner's glands; PG, zone of pyloric 
glands; TZ, transitional zone; FG, zone of fundic glands. 



fibers into laminated sheets at incisura; additional local thickening of circu- 
lar muscle layer within cardiac vestibule opposite incisura cardiaca, possibly 
forming anatomical cardiac sphincter (Fig. 3) (see discussion of muscula- 
ture). Fundic caecum well developed, dilated, expanded superiorly and 
with surface rounded throughout; structurally continuous with midstomach 
(not anatomically distinctive chamber), oriented longitudinally on dorsal 
surface (no dorsal hook). Terminal stomach distal to cardiac vestibule 
long, decreasing gradually in breadth to just beyond midpoint where promi- 
nent sulcus intermedius distinguishes proximal funnel-like portion from 
narrow, relatively long terminal endpiece (2.5 mm in length); terminus 
lateral to sulcus recurved slightly, nearly parallel to cardiac vestibule, uni- 
form in breadth to pyloric sphincter. Pyloric sphincter long and narrow, 
valves of equal length as measured on lesser and greater curvatures; ex- 
panded through medial half to appear narrowly spoon-shaped in cross- 



604 The University Science Bulletin 

section; differing from that of other bats examined in being equivalent in 
depth to circular muscle of pyloric tube that gives rise to sphincter within 
both lesser and greater curvatures; valve aperture narrow. Pyloric junction 
slightly superior to gastroesophageal junction in situ. Musculature generally 
thin throughout much of stomach wall, thickened locally; unusual arrange- 
ment of circular layer in lesser curvature. Rugal folds numerous, reduction 
of mucosal lining throughout relatively large portion of fundic caecum 
and greater curvatures. Lamina propria mucosae extremely thin except 
within fundic caecum and portions of terminal endpiece of terminal stomach. 
(Hands of Brunner moderately abundant. Stomach conforming in most 
features to general configuration common to other obligate insectivorous 
bats. 

Musculature. — Typical three layers observed in both specimens. Longi- 
tudinal layer a simple thin sheet throughout most of stomach, locally thick- 
ened 1) in incisura cardiaca where there is substantial increase in depth 
within the depression, 2) throughout longitudinal zone of fundic caecum 
where it is somewhat thicker than overlying circular layer through much of 
caecal region, and 3) in the mid-region of terminal stomach on greater 
curvature. Circular muscle relatively narrow throughout pyloric tube in 
comparison to other bats examined, uniform in depth throughout length 
of pyloric tube as well as equal in depth in lesser and greater curvatures; 
circular layer locally thickened within broad area in lesser curvature im- 
mediately opposite incisura cardiaca, equivalent in depth to that of incisura 
and extending from midpoint between gastroesophageal junction and sulcus 
intermedius to just beneath gastroesophageal junction, resulting in "ring" 
of thickened musculature at lower end of cardiac vestibule, an anatomically 
plausible cardiac sphincter inferior to gastroesophageal junction although 
clearly not a valvular closure mechanism; circular layer extremely thin 
through fundic caecum and majority of greater curvature, shallower than 
longitudinal layer; bundling of circular muscle occurring only in zone of 
greater curvature proximal to pyloric bend. Muscularis mucosae typical, 
thickest at incisura cardiaca and in fundic caecum. Lamina propria mucosae 
(beneath muscularis mucosae) abundant only in fundic caecum. Depth 
of longitudinal layer 15 to 100 (incisura cardiaca), depth of circular layer 
15 to 140 (incisura cardiaca and opposing wall), depth of mucularis mucosae 
15 to 25. 

Organization of gastric mucosa. —Rugal folds deepest and most abundant 
within tubular portion, originating at pyloric sphincter and radiating into 
mid-stomach, becoming progressively more widely spaced as tube expands; 
frequency of folds constant through tubular portion; least number of folds 
originating along lesser curvature from pyloric bend to cardiac vestibule, 
consistently directed toward lower apex of fundic caecum; upper wall of 



Studies of Stomachs of Selected American Bats 

caecum (lesser curvature) with relatively frequent, short, transversely- 
directed rugae extending downward to mid-region of dorsal and ventral 
surfaces of caecum; folds well developed and longitudinal within cardiac 
vestibule; major portion of greater curvature and all of lateral wall of 
fundic caecum lacking prominent rugal folds (gastric mucosa also distinc- 
tively reduced in depth), which originate in pyloric tube and along lesser 
curvature and terminate in lower middle area of ventral and dorsal sur- 
faces of body of stomach; resulting pattern that of transverse to progressively 
more longitudinally directed rugae from caecum to pyloric tube through 
approximately upper three-fourths of body of stomach and all of tubular 
stomach. All mucosal gland types represented; cardiac and pyloric zones 
relatively short, fundic mucosa occupying remaining surface area except 
for unusually broad transitional zone occurring through nearly all of 
pyloric tube; no Brunner's glands within gastric mucosa. Progressive de- 
crease in mucosal depth from middle fundus through transitional zone 
and pyloric glands to pyloric sphincter with slight secondary increase within 
terminal pyloric mucosa, cardiac glands comparable in depth to transitional 
zone; deepest mucosa within lesser curvature of fundus; glands within 
apex of fundic caecum and in the basal, non-tubular portion of greater 
curvature reduced in length. 

Cardiac glands. — Zone of true cardiac glands extremely narrow; transi- 
tional zone to fundic mucosa slightly longer than cardiac zone; cardiac 
glands short and broad, composed of single type of cell although some local 
variation in morphology of nuclei occurring within gland; cells within 
bulbous base small, elongate, with giant spherical nuclei; upper cells smaller, 
more cuboidal with nuclei decreasing progressively in size although remain- 
ing spherical throughout to epithelium; lumen broad, increasing only 
slightly in breadth toward surface; epithelium in small clumps between 
adjacent glands because luminal openings of glands broad and close together, 
cells elongate, nuclei compressed laterally. Transitional zone to fundic mu- 
cosa with numerous (but scattered) parietal cells throughout length of 
glands. Interglandular and basal lamina propria no more abundant than 
within fundic zone. Depth of cardiac mucosa 160 to 170. 

Fundus.— Most variable in structure of all bats examined; true simple 
tubular glands of branched type only within upper three-fourths of stomach, 
reduced (dwarfed) to thin layers of oxnyctic, chief and mucous neck cells 
within lower fourth (including much of fundic caecum); glands deepest 
between fundo-pyloric bend and medial limit of fundic caecum (inclusive 
of cardiac vestibule), moderate in depth within most of tubular stomach, 
decreasing slightly into transitional zone; reduction of glands within lower 
fundus resulting from general reduction, sometimes complete loss, of one 
or more cell types characteristic of fundic mucosa. Depth of fundic gland: 



606 The University Science Bulletin 




Fig. 4. Argentaffin cell (AC) at base of fundic gland of Rhynchonycteris naso. ZC, zymo- 
genic cell. 

180 to 220 within lesser curvature, cardiac vestibule, part of fundic caecum, 
and zone beneath transitional area within greater curvature; 40 to 100 within 
zone of dwarfed glands. 

Zymogenic cells occupying only lower fifth of gland, most abundantly 
concentrated in stomach within lesser curvature and tubular portion of 
greater curvature, and only scattered infrequent basal elements through most 
of remaining area; zymogenic cells small, pyramidal, largest within ex- 
panded basal arc, decreasing in size and becoming more elongate within 
upper area of intraglandular distribution; nuclei of zymogenic cells small, 
spherical, and juxtaposed to basal membrane. 

Argentaffin cells (Fig. 4) observed infrequently, usually among parietal 
instead of chief cells, approximately same size as mucous neck cells, being 
extremely narrow, triangular, and with nucleus at inner border; point of 
triangle peripheral; nucleus extremely small, spherical, dense argentaffin 
granules lateral (Fig. 4). 

Parietal cells abundant within upper 75% of glands throughout lesser 
curvature from upper midpoint of fundic caecum to pylo-fundic transition; 
also common through small area of lesser curvature just beneath transitional 
zone, extremely sparse to absent through most of greater curvature; parietal 
cells highly variable in size, distinctive intraglandular size gradient with 



Studies of Stomachs of Selected American Bats 607 

increase from bottom to top of distribution, uppermost cells extremely 
large; cells most densely distributed just beneath necks of gastric pits where 
unbroken rows of five to seven often observed, rarely found among chief 
cells; sparsely distributed in greater curvature and generally smaller there 
than in areas of abundance; usually ovoid. 

Mucous neck cells extremely small as compared to adjacent parietal cells; 
rarely occurring singly, usually observed in clumps of four or five, most 
often situated along inner margin of tubule; nucleus half size of parietal 
cell nuclei, oblong, irregular in shape, and with little cytoplasm; mucous 
neck cells composing most of cellular complement within stunted glands 
of greater curvature and fundic caecum in absence of many chief and 
parietal cells. 

Gastric pits shallow through lesser curvature (15-20% length of glands), 
extending to base within dwarfed glands of greater curvature in absence 
of non-mucoid cellular elements; cells lining pits similar to lower mucous 
neck cells, although lumenal mucous border moderate in depth; no promi- 
nent lateral flattening of nuclei within lesser curvature although neck cell 
nuclei compressed within dwarfed glands of lower stomach; surface epi- 
thelium thick, nuclei typically basal, club-shaped, and twice length and 
breadth of gastric pit nuclei; interglandular and basal lamina propria ex- 
tremely sparse. 

Pylo-jitndic transition zone. — Extensive area; parietal cells persisting 
through terminal endpiece nearly to pyloric sphincter, resulting in extremely 
narrow pyloric zone; zymogenic cells gradually reduced in number from 
initial margin of zone, large numbers of mucous cells occurring basally 
with parietal cells throughout zone; loss of parietal cells along terminal 
border of zone also gradual, with persisting cells at aboral margin usually 
basal within glands; increase in depth of gastric pits progressive along 
with loss of parietal elements, beginning in midregion of transition; slight 
reduction in depth from fundic mucosa; depth of transitional mucosa ap- 
proximately 150. 

Pyloric glands. — Pyloric mucosa confined to extremely narrow area at 
pyloric sphincter owing to extensive aboral advancement of parietal cells; 
glands short and broad, equivalent in breadth to fundic glands; cells within 
basal third elongate and narrow, with extremely small, spherical, basal 
nuclei; cells within middle third more globular, nuclei twice size as those 
of basal cells and ovoid, upper cells becoming progressively broader and 
shallower toward surface epithelium with nuclei increasing correspondingly 
in degree of lateral compression; glands relatively uniform in depth and 
breadth throughout distribution, basal fourth of tubule expanded into bulb, 
remaining portion uniform in diameter to surface epithelium; lumen nai 
rowest within bulb, increasing gradually in breadth to surface; surface epi 



608 The University Science Bulletin 

thelium is in small clumps as orifices are broad and close together, cells 
narrow, elongate with similarly shaped nuclei; loose lamina propria more 
abundant beneath pyloric mucosa than fundic, typically sparse between 
glands; depth of pyloric mucosa 150 to 170. 

Glands of Brunner.—Con&ned to short, although broad, massive ring 
occupying only proximal 1.0 mm of duodenum; deepest along distal surface 
of pyloric sphincter, decreasing sharply distally; tubules moderate in breadth, 
cross-sectional morphology identical to that of most other bats examined 
except nuclei extremely small, juxtaposed to basement membrane, and 
perfectly spherical; glands restricted to duodenum, always submucosal. 

Remarks. — The antero-posteriorly expanded tubular stomach of Rhyncho- 
nycteris is well representative of the "insectivorous" condition, with a few 
noteworthy specializations. 

The musculature is substantially better developed on the lesser, as com- 
pared to the greater, curvature, a characteristic unique to this species. This 
condition is greatest at the gastroesophageal junction where local thickening 
occurs, suggestive (although not conclusively indicative) of the presence 
of a non-valvular cardiac closure mechanism. Continuation and expansion 
of the work of Botha (1958, a-c) on gastroesophageal closure mechanisms 
in mammals would be enlightening in this regard. 

The pattern of gathering of the mucosal folds is unique in two respects. 
Many rugae originate along the lesser curvature and are directed somewhat 
downward, toward the lower apex of the fundic caecum, rather than strictly 
longitudinally as in other obligate carnivores. Also, although not unique 
among carnivorous bats (observed to some extent in Plecotus townsendn in 
this study), transversely directed rugae are much in evidence within the 
upper fundic caecum. Secondly, unparalleled reduction of mucosal folds 
and depth of glands occurs throughout much of the greater curvature. 

The gastric mucosa is unique in its impoverished complement of zymo- 
genic cells within the fundic zone (excepting the overall reduction of the 
gastric mucosa in Desmodus). The transitional zone is extensive in terms 
of lengdi and parallels that of Pteronotas in being not only a distinctive 
portion of the mucosa, but by limiting the pyloric mucosa to an extremely 
narrow band at the pyloric sphincter. 

Family NOCTILIONIDAE 

Noctilio labidis (Kerr, 1792) 

Specimens examined (2). — Two specimens from Nicaragua of Noctilio 
labidis minor (KU 105941, 116690). 

Overall dimensions. —Greatest length 12.9-13.5; greatest breadth 7.0-8.0; 
gastroesophageal junction to pyloric sphincter 5.8-6.1; gastroesophageal junc- 
tion to Eundic caecum 4.4-4.7; breadth pylorus at sphincter 2.4-2.6. 



Studies of Stomachs of Selected American Bats 



609 




2mm 



Fig. 5. Mid-longitudinal representation of the stomach of Noctilio labialis; SSE, stratified 
squamous epithelium; CSPH, cardiac sphincter. For explanation of other symbols, see Figure 3. 



General morphology. — Stomach tubular, relatively short and narrow, 
only slightly curved in frontal plane; lesser curvature arched anteriorly; 
anterior and posterior surfaces of stomach convex resulting in egg-shaped 
stomach. Fundic caecum short and narrow, tapering to narrow apex, di- 
rected laterally, not chambered. Esophagus entering stomach at slight angle; 
cardiac vestibule lacking gastric mucosa, stratified squamous epithelium 
extending posteriorly beyond arc of lesser curvature forming lip at entrance 
of esophagus into stomach, and incorporated as lining of well-developed 
cardiac sphincter of mixed type as defined by Botha (1958 a,b); sphincter 
composed of both striated and smooth muscle elements. 

Cardiac sphincter occurring as distinctive fold of stomach wall (Figs. 
5-6) at entrance of esophagus into stomach, asymmetrical, best developed 
on pyloric side; composed of circular muscle layer, muscularis mucosae and 
surrounding lamina propria, and stratified squamous epithelium, but longi- 
tudinal muscle layer not incorporated into fold; circular muscle projecting 
downward to apex of lip and recurving sharply craniad to lesser curvature, 
subdivided into extensive numbers of laminated sheets, nearly uniform in 
breadth throughout junction; muscularis mucosae extensive throughout 
sphincter, fibers separating into radiating bundles at apex of lip, continuing 
as narrow double layer at onset of gastric mucosa; most of sphincter lip 
lined with thin, rugose layer of esophageal epithelium, lateral edge of lip 



610 The University Science Bulletin 

partially lined with gastric glands; loose connective tissue abundant above 
and below muscularis mucosae within that portion of lip lined with squa- 
mous epithelium. Prominent external constriction at junction of abdominal 
esophagus and gastric sac, incisura cardiaca absent, cardiac orifice broad; 
squamous epithelium at cardiac junction and adjacent cardiac mucosa 
similar in depth. 

Terminal stomach distal to cardiac junction moderate in length, curved 
relatively slightly craniad, tapering gradually to sphincter; pyloric bend slight; 
orifice at sphincter narrow because flaps of sphincter long; external pyloric 
constriction prominent, partly owing to extensive expansion laterally within 
duodenum of submucosal glands of Brunner; musculature of pyloric sphinc- 
ter narrow, flaps relatively long, longer and broader on greater than lesser 
curvature; circular muscle layer of duodenum sole muscular contribution to 
sphincter as circular layer discontinuous from pylorus to duodenum with 
pyloric complement ending abruptly at sphincter; sulcus intermedius present 
and well developed approximately 2 mm proximal to pyloric sphincter; 
sulcus angularis occurring midway within terminal stomach. Pyloric and 
cardiac orifices at approximately equivalent levels but not unidirectional. 
Lamina propria mucosae abundant and dense within rugae, less abundant, 
although prominent beneath muscularis mucosae of outer wall. Rugal folds 
extensive in number and depth throughout stomach, unusually prominent 
through portions of greater curvature; gastric mucosa moderate in depth, 
showing familiar pattern of reduction in depth between folds. 

Musculature — Musculature generally deep throughout stomach, best de- 
veloped within most of greater curvature and terminal 3 mm of lesser 
curvature. Longitudinal layer a thin sheet throughout all of lesser curva- 
ture and cardiac junction, abruptly thickened and moderate in depth 
throughout fundic caecum; locally thickened to prominent layer through 
medial 4 mm of greater curvature, again narrowing and remaining narrow 
through terminal portion of greater curvature to pyloric sphincter. Circular 
layer relatively narrow through fundic caecum, locally thickened within 
medial portion of greater curvature (along with underlying layer), again 
narrowing just proximal to sulcus intermedius where local thickening of 
circular layer only is present, narrowing again before increasing sharply 
in breadth to pyloric sphincter; correspondingly thickened zones between 
cardiac junction and pyloric sphincter within lesser curvature, although 
locally thickened areas slightly deeper; gathering of circular fibers into 
bundles extreme in, and restricted to, fundic caecum and greater curvature 
laterally through thickened midregion. Muscularis mucosae typically thin 
layers within stomach wall, slightly thicker within rugae, most prominent 
at cardiac junction. Ganglia of Auerbach's plexus more abundant within 
greater curvature than in lesser curvature. Depth of longitudinal layer 



Studies of Stomachs of Selected American Bats 611 

30 to 60, circular layer 30 to 90 (caecum) to 300 (pyloric tube), muscularis 
mucosae less than 10 up to 20. 

Organization of gastric mucosa. — Mucosa gathered into longitudinal 
parallel ridges from apex of fundic caecum to pyloric sphincter; rugal folds 
best developed of any bat examined; shallow within caecum, increasing 
gradually in depth from origin in fundic caecum through midstomach, 
decreasing again into pylorus; folds gradually decreasing in length and in- 
creasing in depth from lesser curvature posteriad to greater arc. Folds within 
medial half of greater curvature longitudinally directed, relatively short 
finger-like projections of tremendous depth (up to 3.5 mm) directed an- 
teriorly into gastric lumen; few folds originating from lesser curvature (only 
within upper wall of fundic caecum). No transversely oriented rugae; 
moderately developed "zig-zag" configuration in folds within posterior 
three-fourths (absent within region of lesser curvature) from apex of fundic 
caecum to sulcus intermedius, never truly interdigitating; folds shallow, 
infrequent at gastroesophageal junction. Mucosa shallowest in depth at 
gastroesophageal junction, uniform in depth throughout remaining stomach 
excepting slight reduction within interrugal furrows in comparison to sides 
and apexes of rugae; no appreciable reduction within (or at apex of) fundic 
caecum; deepest mucosal glands on sides of rugae within terminal portion 
of transitional zone. 

Zones of gastric mucosa atypical among carnivorous types examined, 
apparently accounting for extremely large transition zone, cardiac glands 
occupying typically narrow band at gastroesophageal junction, equivalent 
in breadth throughout; true fundic mucosa restricted to approximately left 
half, inclusive of region of cardiac junction; transitional zone to pyloric 
mucosa occupying most of terminal half, thus limiting true pyloric glands 
to narrow belt at sphincter; all zones relatively symmetrical regarding length 
of distribution on lesser and greater curvatures (Fig. 5). 

Cardiac glands— Zone of cardiac glands in complete absence of parietal 
cells, extremely narrow; glands short and broad relative to remaining mu- 
cosal elements; single cell type apparently present but extensive morpho- 
logical variation within tubules; cells within basal fourth small and many 
with extremely large, centrally located nuclei; cells in upper three-fourths 
of tubule becoming progressively larger, more elongate, with increasingly 
small, more ovoid nuclei toward upper limit of tubule; tubules not highly 
convoluted, bases usually although not always bulbous and uniform in 
breadth through upper portion; lumen moderate in breadth, slightly broader 
at base than neck; smooth muscle fibers and areolar connective tissue more 
abundant beneath and between tubules than in other mucosal areas; surface 
epithelium thin, cells narrow with large ovoid nuclei located in lower 
middle portion; depth 120 to 140. 



612 The University Science Bulletin 

Fundus. — Relatively restricted fundic zone characterized by marked uni- 
formity in depth and morphology of glands; deepest glands within mid- 
stomach, decreasing only slightly toward apex of fundic caecum, slightly 
reduced in depth within furrows. Consistent zonation of cellular elements 
within fundic glands with little variation in length of zones throughout; 
lower 30% restricted to zymogenic cells with occasional rare argentaffin cell; 
middle 40% occupied by numerous parietal cells with less abundant inter- 
mixed mucous neck cells; upper 30% forming gastric pit; all cellular types 
comparatively small as compared to those of other bats examined. Depth 
of fundic mucosa 150 to 180. 

Zymogenic cells typically pyramidal in form; nuclei spherical, small, and 
usually positioned in middle or lower half of cells, occasionally truly basal; 
basal portion of gland with chief cell complex rarely expanded into bulbous 
structure. Argentaffin cells not observed. 

Parietal cells small with small, slightly ovoid, centrally located nuclei; 
slight variation in cell size apparent but variation mixed, not clinal; parietal 
cells occurring extremely rarely among zymogenic cells, equally uncommon 
within gastric pits of upper portion of tubule, little or no variation in distri- 
bution throughout fundic zone, relatively abundant in view of only mod- 
erate spatial distribution resulting from small size. 

Mucous neck cells among parietal cells compressed horizontally with 
large similarly-shaped nuclei; cells extremely small, infrequent within pa- 
rietal cell zone, never locally abundant within zone. Gastric pits moderate 
in depth, equal in breadth to remainder of gland, hence narrow, and with 
little or no apparent relative increase in depth within lateral arc of fundic 
caecum; pit cells markedly compressed laterally, large with extremely large 
disc-shaped nuclei heavily compressed laterally; cells increasing progressively 
in diameter from neck of pit to surface epithelium. Surface epithelium 
relatively shallow; nuclei large, rounded, and tapered toward base; mucous 
border narrow. Lamina propria mucosae typically scant basally and between 
glands. 

Pylo-fundic transition. — Extensive breadth of transitional area attributable 
not only to pronounced invasion laterally of parietal cells into pylorus, but 
also to relatively abrupt diminution and loss of chief cells considerably 
proximal to pyloric sphincter (zone appearing expanded bidirectionally) ; 
glands essentially like adjacent fundic mucosa, differing in absence of chief 
cells and in that parietal cells distinctively more abundant along with 
numerous mucous neck cells in basal arc; more variation in depth of glands 
and slight increase in depth over that of fundus. Gradual increase in depth 
ot j^istric pits beginning just proximal to sulcus intermedius, zone of in- 
crease occupying approximately terminal third of transitional area, progres- 



Studies of Stomachs of Selected American Bats 

1 «* *-» 



613 







*> J**d$ 




0**$* 



1mm 



6 




»•-» . f 






- 2 . v. 



7 



Fig. 6. Cardiac sphincter at gastroesophageal junction of Noctilio labialis. The fundic 
caecum is toward the left. 

Fig. 7. Pyloric glands of Noctilio labialis. 



614 The University Science Bulletin 

sive increase corresponding to gradual loss of parietal cells within upper 
tubule; lamina propria equivalent to that of fundus. Depth 160 to 250. 

Pyloric glands. — Terminal mucous-type glands (completely lack parietal 
cells) restricted to exceptionally narrow, symmetrical zone extending distally 
slightly beyond lip of sphincter into duodenal lumen; glands moderate and 
slightly variable in depth, slightly broader than fundic glands, remarkably 
uniform in cellular structure (usual single cell type present) ; cells consistent 
in morphology and diameter throughout tubule (Fig. 7), as well as pyloric 
zone, narrow with relatively large, circular, flattened nuclei (twice diameter 
of parietal nuclei) ; cells elongate and with moderately extensive mucous 
border; glands uniform in breadth excepting slight basal dilation, only 
slightly broader than fundic or transition glands. Lamina propria con- 
siderably more abundant basally, slightly more abundant between glands 
than within oxyntic type mucosa. Surface epithelial cells elongate and 
narrow compared to those of fundic glands. Depth of pyloric mucosa 
160 to 200. 

Glands of Brunner. — Restricted to single mass in immediate region of 
pyloric sphincter, relatively abundant; occupying entire distal surface of 
pyloric sphincter and proximal 1.0-1.5 mm of duodenum; thick layer of 
uniform depth, tapering little toward termination of distribution, ending 
aborally at lip of sphincter, not observed within pylorus of stomach; highly 
convoluted tubules of moderate and uniform breadth; cells long and narrow 
with extremely small, spherical, basally located nuclei (Fig. 8). Interposed 
smooth muscle fibers between tubules infrequent compared to other bats 
examined, except Pteronotus parnellii; lumen narrow throughout. 

Remarks. — Noctilio labtalis is unique among bats examined in the occur- 
rence of an apparent anatomical cardiac sphincter at the gastroesophageal 
junction. Botha (1958a) examined the cardiac junctions of 13 species of 
distantly related mammals, describing that of "the bat" as a large, unbroken, 
conical papilla at the cardiac orifice, lined throughout with gastric mucosa, 
the edges of which form an extremely watertight "cork." This description 
is in marked contrast to the situation at the junction in N. labialis, although 
aptly describes cardiac junctions of most other species included in the present 
study. The squamous-lined flaps found in N. labialis also were observed 
macroscopically in one specimen of Noctilio leporinus, and must be con- 
sidered unique to the family Noctilionidae among bats thus far described; 
additional investigations will be required to demonstrate that this structure 
is a physiologically efficient closure mechanism. The specialization at the 
cardiac junction of N. labialis may represent a distinct and independently 
evolved structure, unique among bats to the Noctilionidae and perhaps 
unique within the insectivore-primate-chiropteran line of mammalian evo- 
lution. 



Studies of Stomachs of Selected American Bats 615 



U. i-ki, I 




■v W ff-%*. \ 



Ik. ^ 



> ^W-fc*t> Os. - 1 

Fig. 8. Glands of Brunner of Noctilio labialis. Note the small nuclei. 

Several unique features of the stomach of N. labialis are shared only 
with Pteronotus pamellii. The general outline of the stomach and great 
breadth of the pylorus at the sphincter are similar in these two species, as 
is the extensive pylo-fundic transition zone, with subsequent limiting of 
the pyloric glands to an extremely narrow region at the sphincter. Addi- 
tional features common to N. labialis and P. pamellii include narrow 
oxynctic glands with comparatively small cellular elements, limited distri- 
bution of zymogenic cells within fundic tubules, and lack of distinctive 
reduction of the fundic mucosa within the caecum. 

The stomach of N. labialis clearly follows the basic morphological pat- 
tern exhibited by other carnivorous bats examined, with notable specializa- 
tion of the cardiac junction, perhaps a mechanism to aid in prevention of 
reversed passage of relatively large amounts of food into the abdominal 
esophagus. 

Family PHYLLOSTOMATIDAE 

Subfamily Chilonycterinae 
Pteronotus pamellii (Gray, 1843) 

Specimens examined (1). — One specimen from Nicaragua of Pteronotus 
pamellii juscus (KU 116992). 



616 



The University Science Bulletin 




2mm 



Fig. 9. Mid-longitudinal representation of the stomach of Pteronotus parnellii. For ex- 
planation of symbols, see Figure 3. 



Overall dimensions. — Greatest length S.6; greatest breadth 4.6; gastro- 
esophageal junction to pyloric valve 2.4; gastroesophageal junction to apex 
of fundic caecum 4.0; breadth pylorus at sphincter 1.8. 

General morphology. — Stomach tubular, short relative to breadth, highly 
symmetrical with gastroesophageal junction located approximately midway 
along lesser curvature (Fig. 9) ; upper and lower curvatures nearly parallel, 
but dorsolateral curvature pronounced (dorsal surface convex); fundic 
caecum short, directed slightly anteriorly with no evidence of dilation within 
dorsal surface, apex narrowly rounded; terminal portion of stomach ex- 
tremely short and broad. Gastroesophageal junction and cardiac vestibule 
oriented nearly directly craniad, junction and cardiac orifice extremely 
broad; inferior border of stratified squamous epithelium only half depth 
of adjacent cardiac glands, gastroesophageal junction thus similar morpho- 
logically to that of Natalus; muscularis externa and muscularis mucosae 
continuous between esophagus and cardiac vestibule with no breaks or 
unusual orientation of fibers and no evidence of an anatomical cardiac 
sphincter; cardiac vestibule moderate in length, pattern of gradual coalition 
with lesser curvature similar to that in Plecotus (but vestibule shorter); 
incisura cardiaca absent (but slight thickening of circular muscle layer 
within wall that corresponds to zone of incisura cardiaca in other bats). 



Studies of Stomachs of Selected American Bats 



617 




Fig. 10. Diagrammatic mid-longitudinal illustration of gastroduodenal junction of Pterono- 
tits parnellii. BG, Brunner's glands; a, extension of Brunner's glands beyond lip of pyloric 
sphincter; b, pronounced reduction of pyloric sphincter (lesser curvature); c, small, narrow 
musculature of sphincter within greater curvature; anterior is toward the top in this figure. 



Fundic caecum shallow; mucosa within caecum deep, muscular comple- 
ment of caecal wall thick owing in large part to longitudinal muscle layer 
rather than circular layer. Stomach distal to cardiac vestibule directed 
antero-laterad, tapering gradually to extremely broad pyloric orifice. Muscu- 
lature of pyloric sphincter poorly developed on greater curvature as an ex- 
tremely narrow mass of circular muscle and muscularis mucosae extends 
outward and upward toward duodenal lumen (mass less in length than 
depth of glands of Brunner resting on its distal surface) (Fig. 10); sphincter 
of lesser curvature reduced to simple bulge in circular muscle layer; pyloric 
mucosa lining sphincter equivalent in depth to that of remainder of pylorus; 
pyloric orifice broad and displaced to left within tube; pyloric orifice inferior 
to esophageal junction in situ, although not pronounced because stomach 
nearly horizontal within body cavity. Lamina propria mucosae never 
abundant within stomach wall proper, moderately represented within rugae; 



618 The University Science Bulletin 

gastric mucosa comparatively deep among bats examined, occupying entire 
inner surface and never locally reduced (except extremely slight reduction 
within furrows of pylorus). 

Musculature. — Muscular wall comparatively thick among bats examined 
and best developed among phyllostomatids studied; cardiac vestibule thick 
walled, musculature as deep as that of major portion of lesser curvature 
with circular layer principle component and longitudinal layer simple 
thin sheet. Wall of lesser curvature shallower than corresponding area of 
greater curvature only within terminal stomach, remaining portions (upper 
and lower walls of fundic caecum) approximately equal in depth, being 
slightly thicker than terminal half of lesser curvature; wall of greater curva- 
ture increasing progressively in depth from midpoint to pyloric sphincter, 
becoming two to three times breadth of comparable area within lesser arc 
through terminal 2 mm. Gathering of circular fibers into bundles prominent 
in lower wall of caecum, bundles becoming fewer in number and larger 
through upper wall to base of cardiac vestibule; pattern of depth of longi- 
tudinal layer similar to that observed in Phyllostomus, being exceptionally 
deep through lower and apical wall of caecum, decreasing through lower 
wall of terminal stomach, and extremely thin throughout lesser curvature; 
muscularis mucosae thin double sheet throughout wall, slightly deeper 
within rugae. Ganglia of Auerbacffs plexus numerous except within cardiac 
vestibule and between vestibule and pyloric sphincter along lesser arc; 
lamina propria uniformly scant between muscularis mucosae and tunica 
muscularis throughout entire stomach wall, slightly more abundant beneath 
muscularis mucosae within rugae. Depth longitudinal layer 15 to 45, circular 
layer 30 (fundic caecum) to ISO (pylorus), muscularis mucosae less than 
10 to 15 (within rugae). 

Organization of gastric mucosa. — Rugal folds fewest in absolute number 
as well as least frequent per given area of gastric mucosa among bats 
examined; strictly directed longitudinally throughout entirety of dorsal and 
ventral surfaces of body of stomach from apex of fundic caecum to pyloric 
sphincter; isolated, transversely-oriented folds not observed anywhere within 
stomach lining, and folds not locally reducted in frequency; folds of cardiac 
vestibule continuous upward to gastroesophageal junction, although con- 
forming to general longitudinal orientation (also the case in Plecotus, 
Natalus, Noctilio, and Rhynchonycteris). Depth of mucosa relatively uni- 
form throughout except for cardiac glands; fundic mucosa slightly deeper 
than pyloric mucosa and slightly deeper in lesser curvature than in greater 
curvature, little reduced in fundic caecum; cardiac glands exceptionally 
deep in contrast to sharp decrease in depth of adjacent fundic mucosa; 
cardiac zone similar in depth, breadth, and morphology to that of Phyllo- 
stomus. Combined fundic glands and transitional zone occupying nearly 



Studies of Stomachs of Selected American Bats 619 

all of surface area as pyloric zone extremely short; fundic glands uniform 
in structure within zone and with no secondary regions of noteworthy 
variation from basic plan; transitional zone to pyloric mucosa broad in 
comparison to that of other phyllostomatids and occupies entire lesser curva- 
ture between gastroesophageal junction and pyloric mucosa (also extensive 
along greater arc) ; area of true pyloric mucosa relatively slight and asym- 
metrically distributed on upper and lower surfaces owing to extensive 
transitional zone and asymmetry of pyloric sphincter. 

Cardiac glands. — Zone relatively narrow, equal in length throughout 
distribution; glands similar in morphology to those of Phyllostomus (long 
and narrow) although cellular components not identical; glands nearly 
uniform in breadth throughout, not prominently expanded at base; lumen 
extremely broad and uniform in breadth throughout; cells in lower third 
of tubule larger than those in upper portion, pyramidal, with large elongate 
nuclei compressed against basement membrane; cells in upper two-thirds of 
tubule extremely long and shallow with elongate, prominently compressed 
nuclei; mucous layer of lumen border shallow (in contrast to broad border 
of comparable cells in Phyllostomus). Lamina propria mucosae between 
glands better developed than within remainder of mucosa. Transitional area 
from cardiac to fundic mucosa twice breadth of cardiac zone; gastric pits 
in transitional zone decrease abruptly in depth, glands identical to cardiac 
mucosa except for presence of substantial number of scattered parietal cells; 
surface epithelium equivalent to that of remaining stomach (excepting 
slightly smaller nuclei). Depth of cardiac glands 350 to 390. 

Fundus. — Fundic glands consistently narrow and highly convoluted, 
relatively long; fundic mucosa generally best developed in midportion of 
distribution, decreasing only slightly but progressively in depth in either 
direction laterally; all cellular types comparatively small in relation to other 
bats examined (except Noctilio labialis) by virtue of apparent uniform 
reduction in cytoplasmic mass of cells (because nuclei are universally large 
regardless of cell type); only slight variation in gland morphology within 
zone, variation entirely in number of zymogenic cells and depths of gastric 
pits. Cellular components more or less restricted to specific zones with slight 
alteration in lengths of zones within apical arc of fundic caecum; lower 
15% restricted to zymogenic cells, middle 65% to parietal and mucous 
neck cells and upper 20% to gastric pits. Depth of mucosa 150 to 270 
(cardiac vestibule). 

Zymogenic cells never occupying more than basal 30% of gland; uni- 
form in size, small and globular, not numerous, nuclei extremely large 
relative to cell size and always spherical with nuclei usually displaced 
laterally; zymogenic cells most abundant within glands of cardiac vestibule 
and through area directly posterior to greater curvature; argentaffin cells 



620 The University Science Bulletin 

extremely rare in bases of glands and less than one-half size of chief cells, 
rounded medially, tapering gradually toward periphery (outer limit of 
tubule), nucleus small and spherical. 

Parietal cells, although within extensive spacial distribution, only moder- 
ately abundant owing to large numbers of interposed mucous neck cells; 
parietal cells normally elongate, ovoid, with extremely large, centrally lo- 
cated nuclei; clear cytoplasmic intercellular canals especially numerous; 
pronounced increase in cell size with corresponding decrease in nuclear 
size from bottom to top of distribution, uppermost cells extremely large; 
parietal cells never observed among basal zymogenic cells and reduced 
slightly in frequency within apex of fundic caecum as gastric pits are 
comparatively deeper. 

Mucous neck cells uncommonly abundant among parietal cells as ob- 
served in Phyllostomus, increasing in numbers within gland from origin 
at chief-parietal cell boundary to surface epithelium and often observed 
in large clumps of 5 to 10 cells; mucous neck cells small, irregular in shape 
with large ovoid nuclei usually larger than those of adjacent parietal cells, 
nuclei remain large through gastric pits. Gastric pits shallow throughout 
most of fundic zone, increasing from usual 15% of thickness of tubule to 
20-25% within furrows along apical arc of fundic caecum; neck cells of 
gastric pits with typically large nuclei previously described, nuclei decreasing 
in size only slightly to surface epithelium; mucous border deep throughout, 
lumen both shallow and narrow. Surface epithelium moderate in depth; 
nuclei extremely large, ovoid, not appreciably compressed laterally; lamina 
propria typically sparse basally, poorly developed between rather closely 
packed fundic glands. 

Pylo-fimdic transition zone. — Distribution of transitional glands to pyloric 
mucosa so extensive that true pyloric glands absent in portions of terminal 
pyloric tube; along lesser curvature transitional mucosa occupies all of area 
between inferior border of cardiac mucosa and initial increase in depth 
of musculature near sphincter (see Fig. 9), resulting in extremely broad 
zone; within greater curvature transitional glands extend through approxi- 
mately one-third total length of stomach, essentially all of pyloric tube; 
glands identical in morphology along both curvatures up to terminal margin 
with pyloric glands where junction abrupt on lesser curvature (non-existent 
on greater curvature as true pyloric glands absent); glands essentially identi- 
cal in structure to fundic glands (excepting absence of chief cells) with 
parietal and mucous cells extending to base; deepening of gastric pits abrupt 
at terminal margin of zone on lesser arc and gradual through greater curva- 
ture continuing to duodenal junction (parietal cells persisting within basal 
half of gland to duodenal junction); loss of zymogenic cells abrupt at initial 
margin; surface epithelium identical to that of fundus; depth equivalent 
to that of fundus; depth 210 to 250. 



Studies of Stomachs of Selected American Bats 621 

Pyloric glands. — Extremely limited distribution on lesser curvature; 
glands with parietal cells absent in greater arc and 50% broader than fundic 
or transitional glands; basal third of tubules broadly expanded with large 
cells containing ovoid nuclei moderate in size; cells of upper two-thirds of 
tubule small and more closely resembling mucous neck cells previously 
discussed, and with large, spherical nuclei; glands increasing gradually 
in breadth in upper half of tubule; basal two-thirds of lumen narrow, 
slightly broader in upper third; terminal glands along greater curvature 
generally fitting description given above although much narrower and con- 
taining mixed mucous and parietal cells (considered transitional glands in 
this account); intertubular and basal lamina propria as in fundic and 
transitional zones; depth ISO to 220. 

Glands of Brunner. — Comparatively, best developed among phyllostom- 
atids examined; always submucosal and occur throughout proximal 2.5 
mm of duodenum; in single mass with no isolated lateral clumps; deepest 
along distal surface of pyloric valve, gradually decreasing in depth distally; 
not observed in stomach; typically highly convoluted and branched with 
tubules moderate in breadth; differing from other bats examined (except- 
ing Noctilio) in having extremely small nuclei (always spherical) juxta- 
posed to basement membrane; cells numerous within cross-section of tubule, 
elongate and narrow; glands reduced slightly in breadth toward surface 
in comparison to basal region; pattern of exit into duodenal lumen as in 
other species. 

Remarks. — Simplification of the stomach to a short, symmetrical, tubular 
crescent is characteristic of many obligate carnivorous bats as it is of Ptero- 
notus parnellii. The presence of features observed in insectivorous bats in 
general is more apparent than in Phyllostomus discolor, another insect- 
feeding phyllostomatid. The tunica muscularis is best developed relative 
to stomach size, and the fundic caecum least distinctive, among the seven 
phyllostomatids examined. The terminal portion of the stomach is short 
and broad, totally unlike any other member of the family studied, although 
at least one side of the sphincter is long and narrow as in several other 
leaf -nosed bats. 

General outline, topography, and orientation of orifices are somewhat 
similar to that observed in Noctilio labialis and Plecotus townsendii. The 
most unusual feature is the near loss of true pyloric mucosa at the terminal 
end, with extension of transitional glands laterally. This feature is in contrast 
to the relatively narrow transitional zone in most phyllostomatids. Orienta- 
tion of mucosal folds is strictly longitudinal, a feature also unique among 
phyllostomatids studied. 

Cellular elements of glandular types throughout the gastric mucos 
lend distinctiveness to the stomach by consistently possessing uncommonly 



622 The University Science Bulletin 

large nuclei, in relation to the relatively slight cytoplasmic content of the 
cells. 

Subfamily Phyllostomatinae 

Phyllostomus discolor Wagner, 1843 

Specimens examined (3). — Three specimens from Nicaragua of Phyllo- 
stomus discolor verrucosus (KU 105952-53, 10595S). 

Overall dimensions. — Greatest length 12.1-12.4; greatest breadth 5.0-5.2; 
gastroesophageal junction to pyloric valve 3.5-3.6; gastroesophageal junction 
to apex of fundic caecum 4.1-4.4; breadth of pylorus at sphincter 1.9-2.0. 

General morphology. — Stomach reniform, long and generally symmetri- 
cal; greater and lesser curvatures nearly parallel; marked curvature in 
frontal plane with less marked dorsolateral curvature (dorsal surface con- 
vex) from gastroesophageal junction to apex of fundic caecum; fundic 
caecum long and slightly dilated at apex on dorsal surface; terminal stomach 
long but pyloric endpiece abruptly tapered and relatively short. Cardiac 
vestibule extremely short and tubular, junction with lesser curvature abrupt 
(extends only about 1 mm superior to lesser curvature) ; uniform in diam- 
eter, not tapered, breadth equivalent to that of esophagus; enters body of 
stomach perpendicularly beneath sharply-angled esophagus; thick-walled 
as most of remaining stomach in that musculature well developed. Gastro- 
esophageal junction abrupt; circular and longitudinal layers continuous at 
junction, muscularis mucosae permeated by strands of collagenous tissue 
although most muscle fibers continuous. Incisura cardiaca present, although 
extremely shallow and arises as a result of anterior expansion of the fundic 
caecum, with slight depression resulting at junction of lesser curvature and 
cardiac vestibule; circular muscle layer markedly thickened at incisura 
with compartmentalization of fibers into radiating sheets but no conclusive 
evidence of anatomical cardiac sphincter. Fundic caecum relatively long 
and tubular, directed antero-laterally; dilated at apex and hooking slightly 
dorsally; apex narrowly rounded; caecum not appearing as a distinctive 
chamber but rather as a tubular extension as observed in Plecotus and 
Molossus. Tubular stomach distal to gastroesophageal junction extensive, 
although broader and less extensively recurved than other phyllostomatids 
examined (excepting Pteronotus) ; decreases gradually in breadth to pro- 
nounced constriction (sulcus angularis) approximately 2.5 mm below pyloric 
sphincter (Fig. 11); terminal endpiece to sphincter narrow; gastroesopha- 
geal junction superior to pyloric sphincter in stomach in situ. Pyloric 
sphincter greater in length and breadth on greater curvature and expanded 
along lumenal rim to form narrow lip at orifice; circular layer and muscu- 
laris muscosae form contributing muscle layers; external constriction at 
sphincter slight. Musculature thick with all muscle components well repre- 



Studies of Stomachs of Selected American Bats 



623 




2mm 



Fig. 11. Mid-longitudinal representation of the stomach of Phyllostomns discolor. For ex- 
planation of symbols, see Figure 3. 



sented (including muscularis mucosae) and little reduction anywhere 
except in terminal arc of fundic caecum. 

Rugal folds relatively few in number; gastric mucosa thick throughout, 
reduced in depth only within furrows of fundic caecum, mucosa deepest 
within fundus of lesser curvature in pylorus. Lamina propria best devel- 
oped of all bats examined, particularly within pyloric region and immedi- 
ately adjacent fundic mucosa between muscularis mucosae and circular 
layer, although typically narrow sheets within wall of fundic caecum. 
Stomach generally similar in gross morphology to those of other carnivorous 
species examined. 

Musculature. — Both external layers exceptionally deep throughout (ex- 
cepting major portion of fundic caecum); circular layer substantially deeper 
than longitudinal except in terminal two-thirds of fundic caecum; muscu- 
lature between gastroesophageal junction and lesser curvature moderate in 
depth in comparison to most of remaining stomach, notably deeper than 
musculature of adjacent esophagus and somewhat thickened at junction 
with lesser curvature. 



524 The University Science Bulletin 

Longitudinal layer deep in comparison to other phyllostomatids ex- 
amined; moderate in depth through greater arc of pyloric tube, becoming 
deeper through greater curvature, but decreasing slightly within apex of 
caecum; reduced to extremely thin sheet in upper surface of caecum and 
moderate in depth throughout remainder of lesser curvature. 

Circular layer deep throughout stomach wall (except for apex and upper 
wall of caecum where equivalent in depth to longitudinal layer) ; generally 
deeper throughout greater than lesser curvature; especially pronounced 
local thickening midway along greater curvature of tubular stomach, slight 
local thickening at bases of rugae; circular layer extensively gathered into 
large bundles surrounded by speta and permeated centrally with moderate 
amounts of loose lamina propria originating from normally thin interlayer 
connective tissue sheet; distribution of bundles from midpoint of greater 
curvature to most lateral extremity of fundic caecum. 

Muscularis mucosae uncommonly thick throughout and double-layered; 
substantially thickened in rugal folds as compared to stomach wall proper; 
continuous with muscularis mucosae of esophagus at junction in absence of 
anatomical cardiac sphincter. Connective tissue between circular muscle 
and muscularis mucosae best developed as a continuous layer among bats 
examined; especially pronounced within tubular stomach, but occurring 
as relatively deep region throughout stomach (excluding apex of fundic 
caecum), connective tissue especially well represented within rugal folds; 
collagenous fibers most abundant within rugae, unusually abundant within 
muscularis mucosae but less well represented in outer wall; ganglia of 
Auerbach's plexus abundant. Depth longitudinal layer 15 to 45, circular 
layer 30 (caecum) to 220 (pylorus), muscularis mucosae 15 to 25. 

Organization of gastric mucosa. — Mucosa gathered longitudinally into 
deep, broad rugae throughout all of stomach except along midline of greater 
curvature within non-tubular area, in which rugae are short, discontinuous 
projections that are directed antero-posteriorly (Fig. 12) ; slight "zig-zag" 
orientation of folds within entire non-pyloric stomach, although uniformly 
parallel and never truly interdigitating. Folds relatively few in number 
and reduced locally in frequency within apical arc of fundic caecum; folds 
well developed within short cardiac vestibule, although relatively few in 
number owing to narrowness of tube. Gastric mucosa comparatively deep 
throughout stomach; cardiac glands unusually long; fundic mucosa deepest 
within cardiac vestibule and relatively deep throughout entire distribution, 
also being thick along sides of rugae but slightly shallower at apexes of 
folds and consistently reduced in furrows between folds; average depth 
of fundic glands at apex of caecum slightly less than in remaining fundic 
zone but deeper within lesser curvature compared to greater curvature; 
pyloric mucosa slightly deeper than fundic and somewhat more variable 



<w' 



Studies of Stomachs of Selected American Bats 



625 



K 



■ 










| 0.5 mm , 



I 



*V 






m 



t 4r 








s*# 



ou 









* i.» 



fft 



14 




Fig. 12. A small, isolated, superiorly directed fold of gastric mucosa within the greater 
curvature of Phyllostomtts discolor. 

Fig. 13. Cardiac glands (CG) of Phyllostomiis discolor and gastroesophageal junction, 
showing stratified squamous epithelium (SSE). 

Fig. 14. Zymogenic (chief) cells (ZC) at base of a fundic gland in Phyllostomiis discolor. 

Fig. 15. Gastric pit of fundic gland (Phyllostomiis discolor). Note the broad lumen and 
deep mucous border. 



626 The University Science Bulletin 

in depth. All basic glandular types present; cardiac zone extremely narrow 
(only 0.1 mm at gastroesophageal junction); true fundic mucosa occupies 
all of caecal stomach, much of tubular stomach on greater curvature, with 
less extensive distribution along lesser curvature resulting in correspondingly 
greater representation of pyloric glands along the lesser arc; transition zone 
to pylorus narrow, occupying brief zone just proximal to pyloric constriction, 
equal in length on upper and lower surfaces; pyloric mucosa extensive, 
asymmetrically distributed, occurring throughout short, terminal section 
down through pyloric constriction to abrupt junction with pylo-fundic 
transition zone and cranially to distal surface of pyloric sphincter with no 
invasion into duodenal lumen. 

Cardiac glands. — Extremely thin and relatively straight tubular glands 
(Fig. 13) expanded only slightly at base; gradually increasing in breadth 
within upper half; occupying zone only two to three glands in breadth at 
gastroesophageal junction; single cell type present; cells small, long and 
narrow, with small, laterally compressed nuclei in upper half (Fig. 13) and 
more rounded nuclei in lower half; mucous border narrow; lumen also 
narrow; transitional area to fundus as narrow as cardiac zone and the 
equivalent of three glands in breadth; surface epithelium deep, cells with 
large nuclei that are markedly compressed; depth of mucosa 420 to 460. 

Fundus. — Glands extensively variable within stomach, both in depth and 
in representation of particular cell types. Zymogenic cells in heaviest con- 
centration within midregion of fundic mucosa, decreasing slightly in num- 
ber in either direction laterally; parietal cells also most abundant within 
midregion, decreasing substantially to apex of caecum and increasing in 
direction of pylorus as zymogenic cells decrease; decrease in parietal cells 
in caecum owing to substantial increase in depth of gastric pits throughout 
entire fundic mucosa within fundic caecum, although overall decrease of 
glandular depth slight. Gastric glands relatively short in relation to large 
size of stomach, broad (excepting those of cardiac vestibule and lesser 
curvature immediately surrounding cardiac vestibule). Cell types generally 
restricted to specific zones within gland throughout most of fundic mucosa 
(except mucous neck cells, which occur throughout where zymogenic cells 
absent) ; lower 40% of fundus composed of zymogenic cells with occasional 
scattered parietal cells, middle 30% with parietal cells and frequent mucous 
neck cells, upper 30% represents gastric pits. Depth of mucosa 250 through- 
out most of fundus to 150 within caecum. 

Zymogenic cells (Fig. 14) abundant throughout fundic zone; large, 
irregularly shaped, and of moderate size; nuclei spherical and always laterally 
displaced; cells sometimes polyhedral or ovoid but more regularly triangular; 
lumen narrow through zone of chief cells; base of gland with chief cells, 
although extremely narrow, slightly expanded into bulb. Argentaffin cells 



Studies of Stomachs of Selected American Bats 627 

most abundant within complex of chief cells of any bat examined; extremely 
small in comparison to adjacent chief cells; more or less rectangular with 
large spherical nuclei; outer cytoplasm filled with fine argentaffin granules 
that stain pink with standard haematoxylin and eosin procedure; cells con- 
fined to lower apical arc of chief cell distribution. 

Parietal cells moderate and comparatively uniform in size in relation 
to those of other phyllostomatids examined; most abundant within narrow 
midregion, less abundant as scattered elements among uppermost chief 
cells; usually rounded, rarely ovoid; nucleus always ovoid and moderate in 
size, centrally located within cell; parietal cells may be relatively few due 
to large number of mucous neck cells within midregion of gland. 

Mucous neck cells small, having large, round, peripheral nuclei; mucous 
border deep, much as in Pteronotus; unique among phyllostomatids ex- 
amined in being more abundant in area of parietal cells than parietal cells 
themselves. Gastric pits extremely deep, deepest of any phyllostomatid 
observed, occupying upper 30% through most of fundus to 80% in glands 
at apex of fundic caecum where parietal cells all but absent. Neck cells 
extremely large; mucous border deepest of any bat examined (Fig. 15); 
nuclei large and ovoid in lower half of layer, prominently compressed in 
upper half; lumen a broad chamber increasing only slightly in breadth to 
surface. Surface epithelium deep; cells with extremely large rounded nuclei. 
Lamina propria typically sparse beneath glands. 

Fitndo-pyloric transition. — Zone narrow (see Fig. 11), approximately 
equal in length on both curvatures; loss of parietal cells gradual with few 
cells persisting into basal area of proximal-most pyloric glands; persisting 
parietal cells extremely small; loss of zymogenic cells much more abrupt; 
deepening of gastric pits gradual with progressive loss of parietal and chief 
cells; area of transition in which parietal cells extend to base of gland 
extremely narrow; depth 250 to 310. 

Pyloric glands. — Pyloric glands extensive and reflect generally well- 
developed mucous cell system in Phyllostomus; glands variable in depth, 
local thickenings represent deepest portions of gastric mucosa with glands 
prominently deeper through greater in contrast to lesser curvature; glands 
long and extremely broad (Fig. 16); composed of single cell type (cells 
small with large nuclei in basal half of gland), increasing substantially 
within upper half and becoming giant columnar type with relatively small, 
progressively more ovoid nuclei; lumen extremely broad, hence glands 
uncommonly large; origin of pyloric glands at terminus of fundic mucosa 
fairly abrupt, extending laterally to distal edge of pyloric sphincter; lamina 
propria beneath glands reduced as in remainder of mucosa; pyloric zone 
especially well developed in length and depth in Phyllostomus as compared 
to other phyllostomatids examined; depth of mucosa 200 to 310. 



628 



The University Science Bulletin 




Fig. 16. Pyloric mucosa of Phyllostomits discolor. 



Glands of Brunner. — Glands well developed at pyloric sphincter and as 
moderately deep layer through proximal 3 mm of duodenum; always sub- 
mucosal, never occur within stomach mucosa; distribution along distal 
surface of sphincter only slightly deeper than that within submucosa of 
duodenal wall (former region markedly deepened layer in all other bats 
examined having ! arge complement of Brunner's glands) ; glands occur 
as single mass in duodenum with isolated patches not observed; glands 
virtually identical in cross-sectional morphology to those of Pteronotus; 
orifices between crypts of Lieberkiihn; lumen relatively broad. 

Remarks. — The stomach of Phyllostomns discolor exhibits a mosaic of 
characters of which some are common to other phyllostomatids and others 
are common to carnivorous species irrespective of systematic affiliation. The 
tubular nature of the stomach is consistent with findings for obligate insect 
feeders, and suggests at least a gross morphological adaptation to a diet rich 
in proteinaceous foods. Recent studies (Arata et at., 1967) have alluded to 
the strongly insectivorous feeding habits of Phyllostomus discolor and P. 
hastatus, species belonging to a family predominantly composed of fruit and 
nectar feeders. 

Other features observed in Phyllostomus include a more or less tubular 
undic caecum with only a slight dorsolateral dilation at the apex, short 



Studies of Stomachs of Selected American Bats 629 

tubular cardiac vestibule, well-developed musculature, and lack of pro- 
nounced cranial recurvature of the pylorus. 

Although clearly showing features normally observed in carnivores, 
P. discolor retains conditions common to most other phyllostomatids. These 
include a narrow relatively abrupt pylo-fundic transition zone, dilated fundic 
caecum (although only slightly on dorsal wall), an elongate terminal 
stomach, and a relatively thin muscular complement within the pyloric 
sphincter. 

Mucous production must be extensive owing to the marked development 
of pyloric glands and notably deepened gastric pits of the fundic glands, 
as well as numerous mucous neck cells within the middle portion of the 
fundic glands. 

Subfamily Glossophaginae 
Glossophaga soricina (Pallas, 1766) 

Specimens examined (2). — Two specimens from Nicaragua of Glos- 
sophaga soricina leachii (KU 105984, 105993). 

Overall dimensions. — Greatest length 12.0-13.5; greatest breadth 4.7-4.9; 
greatest length fundic pouch 4.3-4.5; greatest breadth fundic pouch 3.7; 
gastroesophageal junction to pyloric sphincter 4.8-5.3; gastroesophageal junc- 
tion to apex of fundic caecum 5.0-5.2; breadth pylorus at sphincter 1.7. 

General structure. — Large, saccular, bean-shaped, with extensive develop- 
ment of fundic caecum; both saccular and terminal parts well developed 
(Fig. 17) ; marked curvature of stomach in two planes — pronounced curva- 
ture in frontal plane and less pronounced curvature in transverse plane 
(planar orientation of organ in situ). Fundic caecum extensively developed 
(Fig. 18) and set off by sulcus on dorsal surface; pronounced invagination 
of muscularis externa on greater and lesser curvatures resulting in distinc- 
tive "pouch" and semi-ruminate condition; caecum dilated on dorsal side, 
hence caecum spacious; apex blunt; sulcus delimiting fundic "pouch" at 
incisura cardiaca on lesser curvature extending posteriad to midway along 
greater curvature between apex of fundic caecum and pyloric bend (see 
Fig. 17). Region of stomach distal to gastroesophageal junction tapering 
gradually to midway between cardiac and pyloric junctions (where pro- 
nounced crook in tubular stomach occurs, with terminal part recurved 
craniad and terminal end-piece then approximately parallel to esophagus); 
sulcus intermedius present but poorly developed, stomach narrows abruptly 
just distal to "bend" and remains uniformly narrow to pyloric sphincter. 

Incisura cardiaca pronounced and accompanied by noteworthy local 
thickening of muscularis externa; thickened muscle mass dense and partially 
compartmentalized into thin sheets radiating from incisura; sheets separated 
by thin fibrous septa similar to those described by Dearden (1966) in micro- 
tine rodents; cardiac orifice relatively narrow. Cardiac vestibule exti 



630 



The University Science Bulletin 




2mm 



Fig. 17. Mid-longitudinal representation of the stomach of Glossophaga sorkina. For ex- 
planation of symbols, see Figure 3. 



short and with no reduction of included gastric mucosa; longer on pyloric 
than caecal side owing to angular entry of esophagus into stomach. 

Pyloric sphincter asymmetrical (Fig. 19) ; musculature of pyloric sphinc- 
ter substantially more massive on greater than on lesser curvature, muscle 
mass generally thinnest at junction with stomach wall, increasing gradually 
in breadth and terminating at rounded apex on luminal surface; muscle of 
sphincter on greater curvature several times more massive than on lesser 
curvature; relatively shallow; gastric mucosa (pyloric glands) covering 
pyloric sphincter; glands of Brunner completely covering distal edge of 
sphincter muscle mass; only circular and muscularis mucosae layers contrib- 
ute to pyloric sphincter. Gastric mucosa lines entire surface of stomach, 
little reduced anywhere; stomach semi-compartmentalized, possibly adapted 
for storage of food material, presumably to extend length of time during 
which food is exposed to gastric digestion . 

Musculature. — Extensive variation in depth of component muscle layers 
within stomach; outer longitudinal layer always considerably thinner than 
inner circular layer; circular layer thickened at incisura cardiaca and some- 
what compartmentalized into sheets of fibers by thin fibrous septa (re- 
stricted to incisura cardiaca proper). 

Outer longitudinal layer extensively dilated throughout lesser and greater 



Studies of Stomachs of Selected American Bats 



631 



1mm 










Fig. 18. Longitudinal section of the fundic caecum (within dorsal half) oi Glossophaga 
soricina showing fold in body wall (FO) and fundic caecum (FC). 

Fig. 19. Pyloric sphincter of Glossophaga soricina. Note the asymmetry of the sphinctc 
musculature, showing Brunner's glands (BG) and pyloric sphincter (PS). 



632 The University Science Bulletin 

curvatures of fundic stomach distal to cardiac junction; remaining fundus 
(fundic caecum) with relatively thickened longitudinal layer on superior 
and inferior surfaces, whereas circular layer in these two areas only slightly 
thicker than longitudinal muscle and clumped into many fibrous bundles 
along inferior surface; both layers notably dilated in apex of caecum, oc- 
curring there as extremely thin sheets. Pyloric tube musculature extensively 
thickened on greater curvature (both longitudinal and circular layers) (Fig. 
19), narrowing again to a thin remnant just inferior to pyloric valve; muscu- 
lature of lesser curvature of pyloric tube less than one-third breadth of 
corresponding tissue of greater surface at its maximum. 

Muscularis mucosae narrow; filamentous sheets throughout stomach, 
most prominent in pyloric tube. Depth of longitudinal layer 10 to 45, 
circular layer 30 to 140, muscularis mucosae 7. 

Organization of gastric mucosa. — Mucosa gathered longitudinally into 
parallel rows throughout all of pyloric area and tubular fundus up to and 
including cardiac vestibule; folds of saccular fundic caecum oriented longi- 
tudinally on sides (dorsal and ventral surfaces in situ) and radially on 
upper and lower surfaces, as well as apex of caecum and effacing one another 
in spoke-like fashion within caecum; folds broad and deep in tubular 
stomach and reduced somewhat in breadth (although not in frequency) 
in caecum due to general reduction in depth of mucosa in this region; 
folds in mid-longitudinal plane rarely broadly expanded at apex in contrast 
to those of Plecotus; mucosa only slightly deeper at apex of rugae than at 
bases and sides of furrows between rugae; fundic mucosa slightly deeper 
on lesser than greater curvature, reduced in caecum and extreme oral por- 
tion of cardiac vestibule; pyloric mucosa deep in proximal area of pyloric 
zone, reduced substantially in depth at orifices and along pyloric sphincter. 

All basic glandular areas represented; cardiac zone relatively broad 
with transition to fundus abrupt; pyloric glands occupy all of pyloric end- 
piece (that portion distal to point of narrowing) ; transition from fundus to 
pylorus narrow on lesser curvature, twice as broad on greater curvature, 
loss of basal zymogenic cells gradual with deepening of gastric pits and 
establishment of mucous complement of pyloric glands relatively abrupt; 
all of fundic caecum and non-pyloric stomach occupied by fundic type 
glandular mucosa (see Fig. 17 for map of glandular areas). 

Cardiac glands. — Band moderate in breadth (1 mm), extending approxi- 
mately to base of muscular fold of incisura; glands relatively numerous in 
comparison to most other bats examined; demonstrable transition to fundic 
mucosa present, although transition relatively abrupt and characterized 
by scattered parietal cells in basal half of gland; cardiac glands extremely 
short and broad (Fig. 20), of fairly uniform height throughout zone and 
with broad lumens; basal cells small, compact, upper cells larger and 



Studies of Stomachs of Selected American Bats 



633 




?\ 



20 




Fig. 20. Zone of cardiac glands at the gastroesophageal junction of Glossophaga soricina, 
showing cardiac glands (CG), esophagus (ESO), and tunica muscularis (TM). 

Fig. 21. Brunner's glands in the duodenum of Glossophaga soricina. 



634 The University Science Bulletin 

usually with large circular nuclei; gastric pits of cardiac glands broad, in- 
creasing greatly in breadth toward mucosal surface where breadth so great 
as to lend villi-like appearance to glands (Fig. 20) ; loose lamina propria 
extensive between slightly bulbular bases of glands, less extensive with 
onset of fundic mucosa; abrupt transition aborally from stratified squamous 
epithelium of esophagus; depth of mucosa 125. 

Fundus. — Variation in morphology within stomach due to reduction 
in numbers or loss of certain cell types in specific regions; mucosa within 
and adjacent to fundic caecum slightly shallower than that of remaining 
fundus and gastric pits broader, "stunted" condition of glands in caecum 
due to reduction in number of all three cell components (zymogenic, pa- 
rietal, mucous neck) ; gastric pits deep in fundic caecum, narrow and shallow 
in remainder of fundus; lamina propria restricted mostly to narrow belt 
beneath bases of glands. Depth of fundic mucosa 140 to 180 in fundic 
caecum and 170 to 220 in tubular portion of fundus. 

Zymogenic cells usually occupy lower third of gland and occasionally 
basal half in midregion of fundus near gastroesophageal junction, reduced 
or absent only in extreme apex of fundic caecum; rectangular to polyhedral 
with large, spherical, peripherally-displaced nuclei; granulation pronounced 
in specimens examined; parietal cells rarely observed within zymogenic cell 
complex. 

Parietal cells normally in middle third of gland; extremely large, often 
four times breadth of adjacent mucous neck cells and circular to slightly 
ovoid; nuclei large (breadth equals one-third breadth of cell) and circular; 
distinct and marked increase in size of cells from bottom to top of gland 
(cells vary in diameter from 7 microns at base to 18 in upper gland); 
basal cells irregular in shape due to more compact cellular elements than 
within upper portions. 

Mucous neck cells confined to upper two-thirds of tubule, continuing 
into gastric pits (as in all other bats examined) from origin among parietal 
cells; cuboidal or trapezoidal with large spherical nuclei (nearly twice as 
large as those of parietal cells); extending to base of glands in fundic 
caecum along with parietal cells (replace zymogenic cells). Argentaffin cells 
extremely rare, found mostly among zymogenic cells in basal arc. 

Gastric pits of fundic glands distal to caecum shallow, those within 
caecum deeper, broader, 30-60% length of glands; surface epithelial cells 
short and with elongate, laterally-compressed nuclei, decrease markedly in 
size from luminal surface to bottom of gastric pit; mucous border of epi- 
thelium not extensive. 

Pylo-fundic transition. — Extremely narrow zone on lesser curvature, 
more extensive on greater arc, occupies approximate region of pyloric bend; 
depletion of chief cells gradual; broad area with absence of chief cells yet 



Studies of Stomachs of Selected American Bats 635 

with shallow gastric pits; gastric pits deepen abruptly in terminal portion 
of zone; parietal cells numerous and extend into base of glands, relatively 
small throughout glands; depth 200. 

Pyloric glands. — Occupy tubular portion of stomach approximately distal 
to pyloric bend and up to superior surface of pyloric sphincter; glands as 
deep or slightly deeper than those of fundus, although notably broader and 
with substantially thicker, more compact surface epithelium; composed of 
single cell type (with exception of occasional basal argentaffin cell); basal 
mucous cells extremely compact, base of gland bulbular; nuclei large, spheri- 
cal; from midregion upward cells increase in size, nuclei enlarge and be- 
come progressively more laterally flattened and "club-shaped"; lamina 
propria more extensive than in fundic mucosa, areolar connective tissue 
prominent beneath glands; gastric pits broad and lumen of gland extremely 
broad with two glands opening into single pit; pyloric glands well devel- 
oped, extensively distributed when compared to many other bats examined; 
depth of pyloric glands 150 to 250. 

Glands of Brunner. — Limited to proximal-most 2 mm of duodenum 
and never found superior to muscularis mucosae except in limited area at 
apex of pyloric valve; narrow tubules highly coiled with broad lumen; 
cells irregularly shaped with large, spherical, peripheral nuclei (Fig. 21); 
smooth muscle fibers and lamina propria observed only rarely between 
glands. 

Remarks. — The stomach of Glossophaga soricina is most distinctive in 
the presence of a rounded, spacious fundic caecum prominently dilated on 
the dorsal surface, and unique among bats examined with the presence of 
a distinct sulcus deliminating the fundic pouch from the remaining stomach. 
A figure depicting the gastrointestinal tract of Glossophaga included in 
accounts of gross morphology of several bat stomachs by Park and Hall 
(1951) illustrates the dilated portion of the caecum, and corresponds well 
with my findings. The stomach cannot be considered truly ruminate, as no 
rigid partition exists, although the stomach clearly approaches a two-chamber 
condition (Fig. 17). 

Rouk (1968) described and illustrated the stomach of Leptonycteris 
sanborni, a North American nectivore of the subfamily Glossophaginae. The 
stomach of this species agrees well (as illustrated by Rouk) with that of 
G. soricina in general shape and configuration of gross features, particularly 
in orientation and morphology of the terminal stomach, including the 
pyloric sphincter. Also, the orientation and location of the gastroesophageal 
junction as reported by Rouk is virtually identical to that of Glossophaga, 
although the vestibule is somewhat more extensive in L. sanborni. Rouk 
states that the stomach of L. sanborni is "more globular or saccular' 
those of the insectivorous types he examined, although an included : 



636 The University Science Bulletin 

longitudinal drawing does not conclusively illustrate this point. Rouk also 
noted that the pyloric sphincter was thinner than that of insectivorous types 
examined, a feature observed in Glossophaga and several other phyllo- 
stomatids I have examined. Rouk's map of the distribution of glandular 
types shows the glands of Brunner to be extremely restricted in distribution, 
unlike the rather large mass at the pyloric sphincter of Glossophaga. The 
relative depths of the tunica muscularis are similar in the two species, 
although transverse gathering of circular fibers into bundles extends craniad 
more in the fundic caecum of Glossophaga. 

The gastric mucosa of G. soricina is unique among herbivorous phyllo- 
stomatids examined in being rather generalized, composed of relatively 
short, broad glands somewhat resembling those of Plecotus; the three 
cellular components each are more or less confined to one-third the length 
of the tubule and overlap little in distribution. 

Rouk's description of the mucosal folds of Leptonycteris sanborni sug- 
gests existence of some degree of anastomosing, as observed in the fundic 
caecum of Glossophaga, although the description is unclear as to the exact 
location of anastomosing folds, and difficult to interpret comparatively. He 
also reported the presence of what appeared to be clumps of cells identical 
to glands of Brunner within the pylorus. No glands positively identified 
as Brunner's (histologically or histochemically) were observed in the stom- 
ach of Glossophaga. 

Subfamily Carollinae 
Carollia perspicillata (Linneus, 1758) 

Specimens examined (2). — Two specimens from Nicaragua of Carollia 
perspicillata azteca (KU 106036, 106064). 

Overall dimensions.— Greatest length 10.1-11.4; greatest breadth 5.0-5.5; 
gastroesophageal junction to apex of fundic caecum 4.0-4.7; gastroesophageal 
junction to pyloric sphincter 4.1-6.2; breadth of fundic caecum 4.0-4.1; 
breadth pylorus at sphincter 3.9. 

General structure. — Stomach club-shaped and decidedly asymmetrical 
with moderately well-developed cardiac vestibule (Fig. 22) ; tubular stomach 
markedly elongate; marked dorsolateral flexure affecting entire stomach, 
dorsal surface concave; stomach most closely resembling that of Glossophaga 
among bats examined in general stomach outline; gastroesophageal junction 
abrupt, muscularis externa continuous with muscularis of esophagus, muscu- 
laris mucosae thickened slightly although in continuous sheets; no ana- 
tomical cardiac sphincter. Cardiac vestibule moderately developed (Fig. 
23), entering stomach at pronounced angle to lesser curvature creating 
marked incisura cardiaca; expanding gradually from origin at gastroesopha- 
geal junction and disappearing gradually within lesser curvature; extremely 
thin-walled compared to remaining areas of stomach although mucosal 



Studies of Stomachs of Selected American Bats 



637 




2mm 



Fig. 22. Mid-longitudinal representation of the stomach of Carollia perspicillata. For ex- 
planation of symbols, see Figure 3. 



membrane deep within vestibule; incisura cardiaca relatively broad and 
deep (with thickening of circular muscle layer only, no comparable condi- 
tion in other musculature); incisura best expressed within dorsal half of 
stomach as dilated dorsal wall of fundic caecum tends to expand somewhat 
craniad. 

Fundic caecum saccular, appearing as dilated circular bulb lateral to 
vestibule (Fig. 24), heavily dilated and rounded on dorsal (back) surface, 
comparatively small and less distinctive as "chamber" than caecum of 
Glossophaga; caecum distinctive not only externally but by pattern of mu- 
cosal gathering within, never truly separated anatomically by folds of 
musculature or gastric mucosa; musculature of caecum not reduced ap- 
preciably in depth but muscle layers loosely organized with unusually heavy 
complements of elastic fibers within muscularis; complex, oblique orienta- 
tion of longitudinal layer throughout caecum, extensive gathering of circu- 
lar fibers into large numbers of small narrow sheets. 

Sulcus intermedius present, broad and shallow; sulcus angularis rela- 
tively deep, occurring just proximal to pyloric bend; tubular region of 
stomach extremely long and comprising nearly three-fourths total leng 
of the stomach; pyloric bend considerably distant from cardiac junction, 
terminal third of stomach sharply recurved craniad and laterally toward 



638 The University Science Bulletin 

cardiac vestibule; terminal stomach tapering gradually from cardiac vesti- 
bule to sulcus angularis just proximal to bend and remaining narrow to 
pyloric sphincter, with musculature moderate and relatively uniform in depth 
throughout tubular portion. Pyloric sphincter thin, slightly more massive 
on greater than lesser curvature, directed into lumen of duodenum in both 
specimens observed; sphincter composed of circular layer and muscularis 
mucosae; pyloric mucosa lining sphincter deep; external constriction at 
sphincter slight; pyloric tube distal to bend approximately parallel to cardiac 
vestibule, hence direction of gastric orifices approximately equal with both 
directed anteriorly in situ. 

Entire inner surface area of stomach covered with mucosal lining with 
pattern of rugal folds similar to that found in Glossophaga; submucosa 
scanty throughout, included within rugae and best developed within pyloric 
zone. 

Musculature. — Both major muscular layers shallower throughout lesser 
curvature than elsewhere (excepting prominent thickening of wall at 
pyloric bend within lesser arc) ; muscularis externa extremely thin through 
cardiac vestibule along left wall, circular layer the deepest layer throughout; 
muscle wall moderate in depth through lesser curvature to pyloric bend, 
thickened slightly at bend to pyloric sphincter. External layers generally 
equal in depth throughout fundic caecum with circular layer thickened 
slightly at bases of rugae; wall of fundic caecum slightly thicker than that 
of lesser curvature. Longitudinal layer deeper relative to circular layer 
through more of stomach than in any other bat examined; fibers unusually 
oriented and prominently compartmentalized as discussed previously 
throughout fundic caecum, caecum appearing unusually elastic; muscularis 
externa of greater curvature lateral to fundic caecum relatively uniform in 
depth to pyloric sphincter, with no substantial thickening of either external 
layer in pyloric tube and no gathering of circular fibers into bundles in 
lateral two-thirds of tubular stomach. Muscularis mucosae thin, double 
layered, slightly deeper within rugae in comparison to stomach wall proper. 
Depth longitudinal layer 10 (cardiac vestibule) to 80, circular layer 25 
(cardiac vestibule) to 150, muscularis mucosae 10 to 15. 

Organization of gastric mucosa. — Orientation of mucosal folds similar 
to that of Glossophaga but longitudinal orientation better developed along 
ventral and dorsal walls of caecum; folds gathered longitudinally within 
most of tubular stomach and all of cardiac vestibule; folds more or less 
longitudinal throughout upper two-thirds of caecum, with pronounced 
zig-zag orientation and directed transversely into many short folds along 
posterior wall within lower third; dorsal surface of caecum dilated with 
radially arranged rugae similar to those in Glossophaga. Zones of gastric 
mucosa distributed in pattern similar to those of other frugivorous and necti- 



Studies of Stomachs of Selected American Bats 639 

vorous phyllostomatids, but transitional zone relatively long and confined 
to narrow area proximal to pyloric bend on lesser curvature and more 
extensively distributed through pyloric bend on greater curvature; reduc- 
tion in number of parietal cells abrupt within transition, but loss of 
zymogenic cells more gradual, occurring within proximal half; fundic 
zone more extensive on greater than lesser curvature with morphology 
of fundic glands highly variable from one region to another, and unusual 
in several structural features (refer to section on fundic mucosa); extensive 
variation in depth, mucosa slightly shallower within fundic caecum than 
remainder of zone; pyloric zone longer than in Glossophaga, occupying 
all of terminal segment of pyloric tube to rim of pyloric sphincter. Mucosa 
consistently deeper than that of Glossophaga relative to overall size of 
stomach; deepest within tubular portion of fundic mucosa and basal half 
of cardiac vestibule, decreasing only slightly in depth toward lateral ex- 
tremities; mucosa shallowest within furrows between rugae (most apparent 
in caecum) ; mucosa along lesser curvature about same depth as that along 
greater curvature. 

Cardiac glands. — Similar to those of Glossophaga in cellular structure 
and morphology of gastric pits; glands confined to narrower zone than in 
Glossophaga, tubules somewhat narrower, and lamina propria less abundant 
at bases of glands; junction with stratified squamous epithelium abrupt; 
glandular zone consistent in length throughout vestibule; depth approxi- 
mately 90. 

Fundus. — Fundic mucosa characterized by distinctive variability in sur- 
face epithelium and by extensive variability in numbers of various cellular 
components beneath surface epithelium from apex of fundic caecum to 
pyloric mucosa. Depth of fundic mucosa 220 to 370. 

Zymogenic cells in marked abundance only within fundic caecum and 
proximal-most portion of tubular fundus (occupy up to basal 40%, gradu- 
ally reduced in frequency laterally toward pylorus); parietal and mucous 
neck cells increase in abundance toward pylorus with parietal cells increas- 
ing mostly within middle portion of gland and mucous cells progressively 
increasing in abundance throughout entire length of gland; parietal cells 
nearly absent within apical arc of fundic caecum (only three or four cells 
persisting within central area of gland at extreme apex of fundic caecum), 
increasing tremendously from apex of caecum to termination within py- 
lorus (heaviest concentration in mid-stomach and in upper midportion 
of tubule throughout most of fundic mucosa), progressively decrea 
toward base of gland, occasionally scattered among basal chief cells withi 
tubular fundus only. 

Chief cells unusual in structure and arrangement; small and in 
in size; nuclei large, spherical, usually centrally located; cells more or 



640 



The University Science Bulletin 



, 0,5mm , 




Fig. 23. Cardiac vestibule (CV) of Carollia perspicillata. 

Fig. 2-1. Fundic caecum of Carollia perspicillata. 

Fig. 25. Surface epithelium of the fundic mucosa in Carollia perspicillata. 



Studies of Stomachs of Selected American Bats 641 

columnar in shape, inner borders forming narrow but straight lumen (not 
observed in complexes of chief cells of other bats examined); basal portion 
of gland containing only slightly bulbular zymogenic cells. 

Parietal cells exhibiting cline in size within gland, as well as cline in 
abundance within fundic zone; lower cells smallest, increasing in size to 
neck of gastric pit although never truly large relative to tubule breadth; 
parietal cells usually ovoid with large spherical to ovoid nuclei (nuclei 
generally decreasing in size toward lumenal surface). 

Argentaffin cells observed infrequently among chief cells; triangular 
with clear cytoplasm excepting small amount of peripheral granulation 
at one corner of triangle. 

Mucous neck cells substantially more abundant than chief cells in caecum, 
decreasing sharply in abundance toward pylorus (corresponding increase 
in parietal cells, coupled with progressive decrease in depth of gastric pits): 
cellular morphology of mucous neck cells located among parietal cells 
identical to that described for Glossophaga. Gastric pits extremely deep 
and broad within fundic caecum owing to unusually thickened surface- 
epithelium and gastric pit cells; surface epithelium deep throughout stomach 
(Fig. 25), although especially so within caecum; pits 50-60% length ot 
glands at apex, decreasing abruptly within middle portion of caecum, 
shallow (15-20% length of glands) through remainder of fundic mucosa 
and transitional area; pit cells large and cuboidal, increasing in size toward 
surface of mucosa, nuclei enormous and spherical. Surface epithelium com- 
posed of extremely elongate columnar cells in single layer with large, 
laterally compressed nuclei; surface cells decreasing in depth through tubular 
portion of fundus and becoming again thickened within pyloric mucosa. 
Lamina propria beneath glands typically scant; numerous smooth muscle 
fibers between glands within fundic caecum arising from muscularis 
mucosae, and distinctively more abundant than in other bats studied. 

Pylo-jundic transition. — Narrow zone, although relatively broad among 
phyllostomatids studied; extensive on greater arc (parietal cells decrease 
gradually in number toward sphincter with occasional cells lingering through 
pyloric bend); parietal cells that persist in pylo-fundic region not confined 
to specific region within subepithelial portion although usually basal; 
loss of parietal cells more abrupt on lesser arc than greater curvature; 
chief cells disappear abruptly throughout lateral margin of fundic zone, 
gastric pits gradually increase in depth with changes in complements oi 
cellular components; pyloric glands notably decrease in depth (in compari- 
son to fundic mucosa) with reduced depth of terminal transitional mucosa 
reflecting this decrease; surface epithelium deepens abruptly at termir 
margin of zone with no evidence of thickened epithelium within tran- 
sition zone; complement of lamina propria as in fundic mucosa; dept 
ISO to 250. 



642 The University Science Bulletin 

Pyloric glands. — Distribution similar to that found in other frugivorous 
and nectivorous phyllostomatids although somewhat more extensive than 
in Glossophaga; occupying pyloric tube distal to bend; glands short, rela- 
tively straight and tubular, shallowest within inter-rugal furrows. Extensive 
differential gradient in cell morphology from base to surface of mucosa; 
basal cells small with large nuclei, essentially like those within glands of 
Glossophaga and upper cells progressively larger with slight decrease in 
nuclear size and nuclei occasionally compressed laterally. Lumen broad 
and increasing in breadth toward surface; surface epithelium unusually 
thick, comparable to that above fundic mucosa of caecum; unusually ex- 
tensive layer of areolar connective tissue underlying pyloric glands, re- 
stricted to pyloric mucosa and with large number of interspersed mast cells 
and slight increase in interposed collagenous fibers compared to remaining 
mucosa (depth of connective tissue layer approximately 30). Depth of 
pyloric glands 130 to 200. 

Glands of Brunner. — Extremely limited in distribution, occurring as 
single narrow clump on distal surface of pyloric sphincter, and with oc- 
casional scattered glands within mucosa along proximal surface of sphincter 
within pylorus but never within mucosa of stomach wall; glands extremely 
short and broad, highly convoluted; in highest concentration along lip 
of sphincter and decreasing gradually in depth to lateral junction of stomach 
and duodenum; cells large, pyramidal with small ovoid, peripheral nuclei; 
glands most limited in distribution of any bat examined. 

Remarks. — The stomach of Carollia perspicillata somewhat resembles in 
general outline that of Glossophaga soricina. The extent of development of 
the fundic caecum is comparable to that of Glossophaga, although the cae- 
cum is less obviously distinguished from the remaining stomach by a sulcus. 
The distribution and orientation of mucosal folds is similar to that ob- 
served in Glossophaga, but there is somewhat more tendency toward longi- 
tudinal orientation within the dorsal and ventral surfaces of the fundic 
caecum. Carollia and Glossophaga also share the deep surface epithelium 
within the fundic caecum, and a marked variation in distribution of parietal 
cells within the fundic mucosa. 

Carollia is unique in the patterns of distribution of zymogenic and 
parietal cells within the fundic stomach. Chief cells are commonest within 
the fundic caecum (often reduced in this area in other bats) and decrease 
progressively in number laterally, toward the terminal portion. Correspond- 
ingly, parietal cells are nearly absent in the fundic pouch, increasing in 
abundance to the termination of the pylo-fundic transition zone. 

The tendency within the Phyllostomatidae toward elongation of the 
terminal stomach is marked in C. perspicillata. Recurvature of the relatively 
long terminal endpiece is so sharp that it curves back on itself. Speculations 



Studies of Stomachs of Selected American Bats 643 

regarding the possible functional significance of this recurvature follow in 
a later section. 

Structural features at the aboral end of the stomach are strikingly like 
those of Artibeus. The pyloric valve is long and thin, directed into the duo- 
denal lumen in both species. The glands of Brunner differ slightly histo- 
logically but in both species are confined to small masses at the distal 
surface of the pyloric sphincter with few within the submucosa of the 
duodenal wall. 

Subfamily Sturnirinae 
Sturnira lilium (E. Geofrroy-St. Hilaire, 1810) 

Specimens examined (3). — Three specimens of Sturnira lilium parvidens 
from Nicaragua (KU 106086, 106102, 106104). 

Overall dimensions. — Greatest length 9.0-10.2; greatest breadth 6.0-6.3; 
gastroesophageal junction to pyloric valve 4.7-4.9; gastroesophageal junction 
to apex of fundic caecum 4.8; breadth of pylorus at sphincter 1.2-1.3. 

General morphology. — Small, slight curvature along back (dorsal) sur- 
face, hence concave dorsally; fundic caecum saccular; cardiac vestibule 
long and tapering; elongate, narrow pylorus recurved craniad; mucosal folds 
gathered in unusual direction. Gastroesophageal junction markedly su- 
perior to pyloric sphincter and junction with cardiac mucosa abrupt; no 
evidence of anatomical cardiac sphincter; cardiac vestibule (Fig. 26) ex- 
tremely long, thin walled, expanding gradually from gastroesophageal 
junction to terminus in lesser curvature, oriented in S-configuration with 
cardiac orifice directed antero-laterad approximately as pyloric orifice; wall 
of cardiac vestibule more extensive along surface toward pylorus than that 
adjacent to upper caecum. Fundic caecum extremely saccular and thin 
walled; expanded on both upper and lower surfaces into spacious chamber 
with rounded apex; dorsal surface expanded; fold of stomach wall along 
rear (dorsal) surface that distinguishes cardiac vestibule and caecum. In- 
cisure cardiaca deep and narrow; occurring as distinctive depression only 
within posterior half of stomach with slight thickening of circular muscle 
layer at base of incisura, and with orientation of circular muscle fibers 
separated into sheets by septa; prominence of incisura attributable to notable 
anterior dilation of fundic caecum. Tubular stomach distal to fundic caecum 
exceptionally long and narrow; terminal half of tubular region at 90-degree 
angle to remaining stomach and distal-most half with pyloric sphincter 
recurved craniad; pyloric bend gradual; tapering gradually to pyloric bend 
then relatively uniform in breadth to pyloric sphincter. Musculature of 
pyloric sphincter poorly developed throughout, least well represented on 
lesser curvature (reduced to simple "ridge" of circular muscle in mos 
places and occasionally totally absent in portions of its circumfer 
circular muscle and muscularis mucosae contributing to sphincti 



644 



The University Science Bulletin 




2mm 



j 



Fig. 26. Mid-longitudinal representation of the stomach of Sturnira lilium. For explana- 
tion of symbols, .sec Figure 3. 



overlying mucosa deep along muscular ridge, hence total sphincter mass 
extensive; external constriction at sphincter slight; sulcus intermedius pres- 
ent at pyloric bend not prominent. Lamina propria extremely rare every- 
where within stomach wall (except beneath muscularis mucosae within 
pyloric mucosa), prominent within rugae (greatest within terminal portion 
of pyloric tube); rugal folds and gastric mucosa well developed throughout, 
no apparent secondary reduction of either; mucosa especially deep within 
cardiac vestibule. 

Musculature. — Longitudinal and circular layers thin throughout fundic 
caecum and cardiac vestibule, with local thickening of circular layer at bases 
of rugae; gathering of circular fibers into bundles pronounced only within 
lower wall of caecum and occurring to a lesser degree within midregion of 
greater curvature. Circular layer always thicker than longitudinal layer, 
with regions of thickened circular muscle confined to tubular stomach 
(originating on lesser curvature at base of cardiac vestibule, then increasing 
gradually in depth to maximum at pyloric bend, then decreasing slightly 
again to second thickening at pyloric sphincter); essentially identical pattern 
as above within greater curvature, although musculature consistently thicker 



Studies of Stomachs of Selected American Bats 645 

than in lesser curvature in corresponding areas. Longitudinal layer a thin 
sheet throughout stomach, including pyloric tube, not contributing to 
sphincter. Muscularis mucosae a thin double layer throughout with little 
submucosa beneath and most prominent in pyloric area. Ganglia oi Auer- 
bach's plexus abundant only within pyloric tube and greatest within greater 
curvature. Depth longitudinal layer 15 to 20, circular layer 20 to 150, 
muscularis mucosae less than 10. 

Organization of gastric mucosa. — Mucosa gathered uniformly into longi- 
tudinal, parallel folds throughout tubular portion of stomach and cardiac 
vestibule, with slight interdigitation of basal-most rugae in vestibule; 
mucosa of fundic caecum arranged in interdigitating network of deep 
rugal folds with net direction of orientation more or less longitudinal (apex 
of fundic caecum toward pyloric tube), but some secondary transverse 
orientation apparent; interdigitation of folds confined to saccular fundus 
proper, fundic mucosa of mid-stomach gathered longitudinally with little 
or no gathering into short discontinuous folds along arc of greater curva- 
ture; remaining mucosa consistently gathered on long, continuous ridges 
traversing greater part of length of stomach. 

Fundic mucosa uniformly shallow on apexes of rugae within fundic 
caecum and mid-stomach, no local reduction within apex of caecum and 
notably deepest throughout cardiac vestibule where glands are consistently 
longest and narrowest; pyloric glands notably shallower than fundic or 
cardiac mucosa, although mucosal membrane of pyloric tube exceptionally 
thick owing to underlying layer of glands of Brunner; both fundic and 
pyloric mucosa thicker on lesser curvature than greater. 

Four glandular types recognized within mucosal membrane including 
glands of Brunner (submucosal in duodenum) beneath certain portions; 
cardiac glands confined within typically narrow belt at gastroesophageal 
junction, zone equal in length throughout; fundic mucosa occupying all 
of cardiac vestibule and saccular fundus, and mid-stomach up to depletion 
of chief cells proximal to pyloric bend; transitional area narrow and un- 
equally distributed on lower surfaces, confined to narrow region within 
pyloric bend; pyloric mucosa occupying entire surfaces of short pyloric 
tube distal to pyloric bend and extending slightly beyond pyloric sphincter 
into duodenum; glands of Brunner present in typical distribution at superior 
end of duodenum but also occurring within gastric mucosa beneath entire 
pyloric gland complement and most of pylo-fundic transition area, gradt 
ally decreasing in abundance from gastroesophageal junction througl 
pylorus to termination within pyloric bend; no intergraduation of 
Brunner's glands with zymogenic cells of fundus, hence junction abrupt. 

Cardiac glands.— Short, thin and tubular within narrow zone at ga 
esophageal junction; junction with stratified squamous epithelium 



646 The University Science Bulletin 

glands narrow and composed of single cell type; glands two to three times 
breadth of adjacent fundic glands; mucous cells extremely large (twice size 
of surface epithelial cells of remaining stomach); cells pyramidal with 
small, ovoid nuclei within lower half of cell (not juxtaposition to basal 
membrane), cells increasing in size from base to surface; mucous border 
deep; lumen extremely broad and expanded at base along with gland; 
transitional zone to fundic mucosa absent; glands distinctive in morphology, 
easily distinguished from remainder of gastric mucosa; depth 60, breadth 
of zone 30. 

Fundus. — Glands extend throughout all of saccular and tubular stomach 
up to pyloric bend; glands uniform in basic morphological pattern, but 
varying slightly in depth and breadth in cardiac vestibule in comparison to 
remainder of fundic stomach. Glands generally short and broad with ex- 
tremely shallow gastric pits; chief cells extensively distributed within gland, 
succeeded by thin but distinctive zone of primarily mucous neck cells 
superior to chief cell complex, with parietal cells confined mostly to upper 
half and most prominent immediately beneath gastric pits; no distinctive 
reductions in depth of mucosa anywhere within mucosa including fundic 
caecum; mucosa of cardiac vestibule considerably deeper than remainder of 
fundus, owing to substantial increase in relative numbers of parietal cells, 
and also gastric pits somewhat lengthened. Depth of fundic mucosa 120 to 
170 within tubular stomach and fundic caecum and up to 220 in cardiac 
vestibule. 

Basal zymogenic cells extensive in numbers within glands and generally 
throughout fundic region, may occupy up to 60% length of gland (normally 
basal 40-50%); zymogenic cells large, polyhedral, irregular in size, little 
apparent size gradient within gland; nuclei small and ovoid to spherical, 
peripheral although not necessarily juxtaposed to basal membrane; large 
and extensive chief cells result in notable basal expansion of fundic glands, 
chief cell complex most conspicuous element within glands. Cells strictly 
definable as argentaffin cells not observed but presumed present. 

Mucous neck cells particularly abundant within medial 20% of gland, 
large and with small peripheral, spherical nuclei; clear, relatively extensive 
cytoplasm; upper mucous neck cells among parietal cells smaller with larger 
nuclei and wedged between adjacent parietal cells. 

Parietal cells long, narrow, ovoid and with oval nuclei; nuclei vary con- 
siderably in size and often displaced peripherally; size gradient within 
cell distribution present but not pronounced; cells confined mostly to upper 
half of gland and increase progressively in frequency to maximum directly 
beneath gastric pits; parietal cells occurring infrequently among zymogenic 
cells; relatively uniform in numbers throughout fundic mucosa although 
somewhat more abundant within cardiac vestibule than remainder of area. 



Studies of Stomachs of Selected American Bats 647 

Gastric pits shallow, although somewhat variable in depth, usually oc- 
cupying upper 10-15% total length of gland; cells small with large con- 
sistently spherical nuclei throughout pit (no lateral compression); lumen 
narrow; surface epithelium shallow and cells small with spherical nuclei, 
mucous border not deep. 

Pylo-fundic transition zone. — Narrow zone at pyloric bend, more ex- 
tensive on greater than lesser curvature; unusual in having glandular portion 
in form of cellular elements identical to glands of Brunner; transitional 
glands generally intermediate in height between fundic and total pyloric 
mucosa (inclusive of basal cells of Brunner's glands); depletion of zymo- 
genic cells abrupt at beginning of zone, with parietal cells persisting through- 
out as small remnants confined to area above basal mucous elements and 
decreasing in number gradually toward pyloric sphincter; parietal cells 
not observed among basal mucoid cells (presumed glands of Brunner) with 
no region of gastric mucosa existing in which parietal cells extend to base; 
gastric pits increase gradually in depth with progressive loss of parietal 
cells; surface epithelium identical to that of pyloric area, shallow, with 
small cells having small and spherical nuclei; depth 180 to 200. 

Pyloric glands. — True pyloric mucosa restricted to short pyloric tube 
distal to pyloric bend; glands stunted owing to layer of underlying glands 
of Brunner; narrowly expanded into bulb at base, sharply increasing in 
breadth from midregion to surface epithelium, resulting in funnel-shaped 
appearance; depth variable, generally increasing gradually from origin 
within pyloric bend to termination at pyloric sphincter; composed of single 
cell type; cells small and with extremely large spherical nuclei which de- 
crease slightly in size to surface epithelium and never are laterally flattened; 
cells narrow and elongate with extensive mucous border; lumen broad, 
increasing gradually toward surface; thin layers of connective tissue and 
smooth muscle fibers derived from muscularis mucosae separating pyloric 
glands from underlying glands of Brunner; surface epithelium relatively 
thick when compared to that of adjacent transitional glands; lamina propria 
thin layer beneath pyloric glands; basal glands of Brunner discussed in 
following section; pyloric glands do not penetrate into lumen of duodenum; 
depth 100 to 240 (not including basal glands of Brunner). 

Glands of Brunner.— Moderate in abundance within duodenum, occu- 
pying proximal 1.5 mm and occurring in single mass continuous with com- 
plement in stomach; heaviest concentration at sphincter; tubules decreasing 
progressively in length toward termination in duodenum; submucos 
within duodenum. 

Glands occupy basal position within mucosal membrane of pylor 
to lesser extent, pylo-fundic transition area (Figs. 26, '. 
and submucosal within pylorus (beneath muscularis mucosae) 



648 



The University Science Bulletin 



0.1mm 






^** %<; 



* 

* 












^ 



*V 



>* 



'^-v* 
















''/ r 




Fig. 27. Glands of Brunner (BG) within the pyloric mucosa (PM) of Sturnira lilium; the 
pyloric sphincter (PS) also is seen. 

Fig. 28. Glands of Brunner (BG) within the pylorus of Sturnira lilium showing reaction 
to I'AS; the pyloric constriction (PC) marks the limit of the pyloric stomach (left of con- 
striction). 



Studies of Stomachs of Selected American Bats 649 

usually observed superior to muscularis mucosae (Fig. 28); increase pro 
gressively in depth from origin at pyloric bend to distribution within duo- 
denum. 

Morphology of glands identical throughout (Fig. 27); tubules highly 
convoluted, branching into two or three units and consistently large through- 
out entire distribution; tubules open lumenally by way of pores between 
adjacent pyloric glands or crypts of Lieberkuhn depending upon location 
within gastrointestinal tract; tubules never an anatomical subunit of any 
other glandular type; cellular morphology of Brunner's glands consistent 
throughout distribution; cells large, pyramidal, with light granulation; 
nuclei extremely small, juxtaposed to basement membrane, moderately to 
heavily compressed laterally; glands large in comparison to most other 
bats examined and with expansive lumen; muscularis mucosae and lamina 
propria comparatively well developed beneath pyloric complement of Brun- 
ner's glands. 

Remarks. — Sturnira demonstrates the most pronounced invasion of the 
glands of Brunner into the pyloric stomach of any bat examined. Although 
this condition is not restricted to this species, it is most obvious and best 
developed here. It is interesting to note that in addition to having large 
numbers of Brunner's glands beneath the entire distribution of pyloric 
mucosa, the pyloric sphincter is distinctively reduced, and indeed virtually 
absent in portions of the gastroduodenal junction. This condition suggests 
that the invasion orally of Brunner's glands may be partially a developmental 
by-product, in the absence of a prominent barrier conceivably produced 
by a muscular pyloric sphincter. Although Molossus, with its prominent 
pyloric sphincter, also apparently has Brunner's glands beneath the pyloric 
mucosa, the condition is qualitatively and quantitatively much less distinctive 
than in Sturnira. Investigations of gastrointestinal development in bats 
should prove useful in determining any relationship between the size of 
the pyloric sphincter and the extent of Brunner's glands. To my knowledge, 
the studies herein reported include the first mention of a distinctive comp- 
plement of the glands of Brunner in the stomach of a placental mammal. 

The musculature of the stomach wall of Sturnira is similar to that 
observed in other non-carnivorous phyllostomatids— thin and with the 
circular layer arranged locally into bundles enclosed by thin, tissue-likc 
septa. 

Subfamily Stenoderminae 
Artibeus lituratus (Olfers, 1818) 
Specimens examined (3).— One specimen from Chiapas (uncatalogi 
and two from Nicaragua of Artibeus lituratus palmarum (KU 10i 

106236). 

Overall dimensions.— Greatest length 11.6-13.5; greatest breadt 



650 



The University Science Bulletin 




2mm 



Fig. 29. Mid-longitudinal representation of the stomach of Artibeus lituratus. For explana- 
tion of symbols, see Figure 3. 



greatest breadth of fundic caecum 5.5-6.0; gastroesophageal junction to py- 
loric valve 6.0-6.1; gastroesophageal junction to apex of fundic caecum 
8.4-9.1; breadth of pylorus at sphincter 1.6-1.7. 

General morphology. — Similar in many gross morphological features to 
stomach of Sturnira lilium, although considerably larger and extremely 
saccular, with unusually large fundic caecum and long, broad cardiac 
vestibule; dorsal surface concave and with marked dorsolateral curvature 
of entire stomach. 

Pyloric tube relatively long and narrow (Fig. 29), recurved sharply 
anteriorly distal to pyloric bend; musculature thick; pronounced sulcus 
intermedius in central portion of tube; marked external pyloric constriction 
at sphincter; nearly entire internal surface of terminal endpiece lined with 



Studies of Stomachs of Selected American Bats 

ANTERIOR 



651 



RIGHT 




Fig. 30. Diagram showing anastomosing rugal folds with extensive interdigitation of 
ridges in the fundic stomach of Artibeus lituratus. Note the anterior-posterior, rather than 
longitudinal (left to right) direction of the folds. 



pyloric-type mucosa; rugae (Fig. 30) and inclusive muscularis mucosae and 
lamina propria unusually well developed along greater arc of pyloric tube 
beneath pyloric glands, underlying submucosa unusually well developed 
in pylorus; phyloric sphincter (Fig. 31) of circular muscle and muscularis 
mucosae with slight contribution of longitudinal muscle layer at extreme 
base of sphincter, gastric mucosa well developed on musculature of valve, 
sphincter muscle broad at duodenal lumen in preserved specimens, sphincter 
muscle broad at origin on stomach and tapering to narrow endpiece me- 
dially, sphincter symmetrical; sphincter funnel-shaped rather than disc- 
shaped and must produce effective seal upon contraction of circular muscle 
if functional; pyloric orifice displaced considerably inferior (4-5 mm) to 
cardiac orifice with both directed in nearly same direction. 

Incisura cardiaca present but poorly developed. Cardiac vestibule (Fig. 
32) massively developed in all directions into distinct chamber (externally) 
originating at narrow gastroesophageal junction and increasing gradually 
in breadth to lower terminus within lesser curvature; vestibule in one spec 
men with slight S-configuration, anteriormost region reflected laterally and 
parallel to pyloric tube (probably result of slight "tilt" of stomach in sit 



652 The University Science Bulletin 

vestibule expanded frontally yielding greater external as well as internal 
distinction to region than in Sturnira; cardiac orifice substantially superior 
to pyloric orifice. 

Vestibule emptying into greatly dilated saccular fundus divisible into- 
two general areas, 1) central portion between pyloric bend and fundic 
caecum and 2) extensively dilated fundic pouch; fundic caecum extremely 
large, more rounded on dorsal surface {in situ) than in Sturnira; tapering 
laterally to narrow apex, distinguished from remaining fundus by fold in 
stomach wall on dorsal surface as in Sturnira; gastric mucosa not decreasing 
in depth within caecum, thin walled. 

Lamina propria somewhat more abundant than in most other bats 
examined but non-muscular submucosa generally thin throughout (with 
exception of pyloric tube, and within and at base of rugae in fundic caecum, 
where loose connective tissue often dense and moderately abundant). 
Stomach clearly adapted for acceptance of, and perhaps storage of, large 
amounts of food. 

Musculature. — Muscle layers of fundic caecum and fundus of mid- 
stomach irregular in depth and morphology as in Sturnira; stomach wall 
extremely thin throughout most of fundic stomach and cardiac ampulla 
(except for circular layer at origin of rugal folds where internal muscle layer 
notably thickened and grouped into large bundles surrounded by thin septa 
of connective tissue); musculature especially thin in cardiac vestibule; circu- 
lar muscle also thickened at base of rugae, single large blood vessel ob- 
served within lamina propria at base of, and ascending into, each rugal 
fold with system best developed in fundus, to lesser degree in pyloric tube. 

Longitudinal muscle layer uniformly thin throughout most of stomach, 
slightly thickened in pyloric tube in comparison to remaining stomach. 

Circular muscle greatly thickened thoughout greater curvature of pyloric 
tube, decreasing gradually in depth through pyloric bend; circular muscle 
of lesser wall of pylorus thickened locally proximal to pyloric sphincter, 
thin throughout remainder of pylorus; lamina propria above circular layer 
relatively thick within greater arc of pylorus. 

Muscularis mucosae forming generally thin sheet beneath gastric mucosa 
but thicker beneath gastric mucosa and within rugal folds of pyloric tube. 
Musculature of gastroesophageal junction thin, lacking local thickening 
except for definite bundling of circular layer; no conclusive evidence sug- 
gesting presence of active cardiac sphincter, although local segregation of 
circular muscle fibers reveals presence of questionable morphologic speciali- 
zation. Depth longitudinal layer 20 to 30, circular layer 20 in fundic caecum 
to 1S5 in pyloric tube, muscularis mucosae 10. 

Organization of gastric mucosa.— Mucosa arranged throughout fundic 
stomach into large, complex, interdigitating networks of folds (more com- 



Studies of Stomachs of Selected American Bats 



653 




Fig. 31. Pyloric sphincter (PS) of Artibeus lituratus. 
Fig. 32. Cardiac vestibule of Artibeus lituratus showing fold in body wal 
denum (DUO), and cardiac vestibule (CV). Anterior is to the left in this photograph. 



654 The University Science Bulletin 

plex than in Sturnira) with more or less regular pattern as illustrated in 
Figures 30, 32; general direction of orientation antero-posterior, angled 
such that folds progress antero-posteriorly, but directed upward and to the 
right; interdigitation more pronounced in fundic caecum than remainder 
of fundus; folds in pyloric tube irregular in shape, clearly directed more 
longitudinally than those of fundus and not interdigitating. Mucosal folds 
generally extremely well developed without local reductions anywhere in 
fundus, although slightly shallower pyloric glands on musculature of pyloric 
sphincter than elsewhere; mucosa unusually uniform in depth throughout 
fundus, including furrows between folds, less uniform in pyloric tube 
owing to thickening on sides of rugae and reduction in depth within fur- 
rows between adjacent folds. 

All glandular types present, although pylo-fundic transitional glands 
reduced to remnant; cardiac glands occupying narrow belt at gastroesopha- 
geal junction; fundic glands occupying all of remaining surface area except 
pyloric tube; fundic glands long and narrow, exhibiting notable specializa- 
tion in distribution of cellular types; pylo-fundic transition zone extremely 
narrow, basal zymogenic cells of fundus extending to pyloric bend; pyloric 
glands extensive, unusual in appearing to consist of two cellular types, form- 
ing mass of enlarged pyloric sphincter. 

Cardiac glands. — Narrow zone inferior to gastroesophageal junction; 
similar to pyloric glands in basic morphology, but only single cell type 
present; cardiac glands longer and broader than adjacent fundic glands, 
broadly expanded basally, cells long and narrow in single dense layer; 
nuclei spherical with marked size gradient of decrease from lower to upper 
part of glands; luminal mucous border extensive (over twice breadth of 
nucleus) ; broad foveolae plus extensive mucous border yielding relatively 
broad glands; junction with fundic mucosa abrupt; transitional element 
absent as zymogenic cells present within glands adjacent to lower cardiac 
elements; junction with esophageal epithelium equally abrupt; lamina 
propria beneath glands thin; depth 210 to 280. 

Fundus.— Distribution of glands similar to that of Sturnira; mucosa deep 
and of rather uniform depth throughout stomach; extends from posterior 
limit of cardiac zone through cardiac vestibule and fundic caecum laterally to 
pyloric bend; arrangement of mucosal folds complex; glands long and ex- 
tremely narrow, tightly compact, little or no apparent convolution, expanded 
at upper and lower ends but narrowing in midregion. Depth of fundic 
mucosa 160 to ISO. 

Zymogenic cells small, pyramidal with small, spherical nuclei (usually 
in outer periphery of cell but occasionally displaced medially); cells rela- 
tively limited in distribution within basal 15-20% of gland, never in complete 



Studies of Stomachs of Selected American Bats 655 

absence of parietal cells; uniform in distribution within all glands containing 
parietal cells. Argentaffin cells not observed but assumed present. 

Parietal cells remarkably abundant, occurring throughout virtually en- 
tire length of gland to just beneath surface epithelium, most heavily con- 
centrated in medial 60% where abundance virtually excludes mucous neck 
cells; cells unique among bats examined in being abundant among chief 
cells at base of glands (although less abundant than in medial portion) 
and somewhat less common immediately beneath gastric pits; cells generally 
large, circular, uppermost cells often four times breadth of those at base, 
occasionally as broad as gland itself; nuclei also large, usually ovoid, most 
often peripheral with nuclei of basal-most cells extremely large and ovoid. 

Mucous neck cells extremely small and sparsely distributed among 
parietal cells; less frequent than parietal cells in medial half of gland where 
usually located along inner surface of tubule; nuclei spherical and smaller 
than those of parietal cells. 

Gastric pits extremely short and thick-walled, up to 15% length of 
gland but usually less; cells lining pits large, densely packed, with large 
nuclei and lumen narrow; surface epithelium extremely thick, represented 
by dense clumps of epithelial cells with large, club-shaped nuclei. 

Pylo-jundic transition zone. — Region so narrow as to appear nearly 
absent; replacement of parietal cells by mucous neck cells abrupt in pyloric 
bend; zymogenic cells present in base of glands in proximity of pyloric mu- 
cosa, although reduced in abundance; gastric pits abruptly deepened at 
terminal border of transition zone; depth about 170. 

Pyloric glands. — Extensive and well-developed area occupying most of 
tubular stomach distal to pyloric bend and terminating aborally at apex 
of lip of pyloric sphincter; differing distinctively in morphology and dimen- 
sions from remaining gastric mucosa including cardiac glands; pyloric glands 
consisting of two apparent cellular types; approximately lower third occu- 
pied by small light-colored cells with small round nuclei and slight cyto- 
plasmic granulation; lower portion of gland bulbular; cells of upper two- 
thirds larger and more ovoid, mucous border twice as broad as in lower 
cells; basal cells clearly unique within mucosa and notably similar to cells 
of Brunner's glands; pyloric glands long and broad, nearly twice as broad 
as cardiac glands; surface epithelium thicker than that of fundus, nuclei 
laterally flattened, mucous border extensive. Lamina propria unusuall) 
developed along with submucosal elements in general, especially 
muscularis mucosae thick in rugae as well as stomach wall prope 
submucosa relatively abundant beneath muscularis mucosae, 
frequent above; extensive deposits of adipose tissue occurring 
of pyloric tube (beneath muscularis mucosae) and to lesser ex 
stomach; connective tissue immediately beneath pyloric gl 



656 The University Science Bulletin 

tensive than in fundus. Glands more irregular in depth than those of fundic 
glands and reduced in depth within furrows between rugae and along 
surface of pyloric sphincter. Depth of pyloric mucosa 140 to 185. 

Glands of Br unner. — Restricted to immediate vicinity of pyloric sphinc- 
ter and fewest in number of any bat examined (excepting Carollia) ; occur- 
ring in narrow belt along distal surface of sphincter but not extending be- 
yond lip of sphincter; large (more than twice breadth of pyloric gland), 
compound tubular, not highly convoluted; cells long and narrow with 
diffuse granulation, more like mucous cells of gastric mucosa than in other 
bats examined; nuclei small, spherical, basal, never laterally compressed; 
lumen large, as broad as single gastric gland, empties between crypts of 
Lieberkuhn; glands unique in morphology among bats examined. 

Remarks. — The stomach of Artibeus lituratus clearly is adapted to accept 
large quantities of food material and may well serve to retard, secondarily, 
its transport through the digestive system. Extensive development of the 
cardiac vestibule and fundic caecum can only suggest modifications to 
assimilate a bulky diet, probably normally fruit of trees and shrubs (Good- 
win and Greenhall, 1961). Thin musculature of the entire non-pyloric 
stomach would allow for expansion, and thus for consumption of large 
quantities of food. 

Several specializations occur that probably are related to retarding 
gastric flow through the stomach, a phenomenon not uncommon among 
mammals that feed primarily on herbaceous material. The orientation of 
mucosal folds (Fig. 30), for example, would provide a mechanism for di- 
recting some food material laterad into the unusually spacious fundic cae- 
cum, upon contraction of the stomach wall, with the thickened circular 
muscle bands within the bases of the rugal folds providing the necessary 
contraction. Unusually extensive development of the pyloric sphincter and 
tube may be modifications providing an effective closing mechanism against 
large quantities of food, although the importance of an anatomical pyloric 
sphincter in this regard may be in question in mammals (Selkurt, 1966). 
The pyloric tube is extremely narrow in relation to total stomach size and 
probably offers resistance to an escaping bolus. 

The fundic mucosa exhibits reduction in numbers of zymogenic cells 
within individual glands, although presence of these cells is uniform 
throughout the entire region. Inasmuch as chief cells are the suspected site 
of secretion of pepsin, it is not surprising that they are less abundant in this 
primarily fruit-eating bat than in most other species studied. However, the 
unusual abundance of HCl-producing parietal cells is not clear. In addition 
to its function of providing the proper pH for activity of gastric proteases, 
hydrocloric acid also functions in prevention of bacterial growth and stimu- 
lation of secretion of secretin and pancreozymin in the duodenum, as well 



Studies of Stomachs of Selected American Bats 657 

as enzymes by the pancreas (Glass, 1968). Perhaps the extensive parietal cell 
complement in Artibeus can be linked to one or more of these secondary 
functions. No other species of bat examined exhibited such extensive num- 
bers of parietal cells and, for the present, this feature must be considered 
unique in Artibeus. 

Although unique with respect to certain features, the stomach of 
Artibeus is similar to that of other phyllostomatids in gross morphology. 
Artibeus resembles phyllostomatids generally in possession of a saccular, 
asymmetrically formed, muscular, and recurved pylorus, and thin muscu- 
lature throughout a well-developed globular fundic caecum. 

Although distinctive, the stomach of Artibeus lituratus is clearly similar 
topographically to that of Sturnira lilium. Similarities include the broad, 
elongate, slightly S-shaped cardiac vestibule, well-developed saccular fundic 
caecum, short and sharply recurved terminal endpiece, and (internally) the 
interdigitating, diagonally-oriented mucosal folds. Differences include a 
broad pylo-fundic transition zone and predominance of chief cells in the 
fundic mucosa in Sturnira, in contrast to an extremely narrow transitional 
element and persistence of parietal cells as the most conspicuous cell type 
of the fundic mucosa of Artibeus. 

It is of some interest to note that Grasse (1955), in a classification of 
stomachs of bats into 1) megachiropteran, 2) microchiropteran and 3) 
Desmodus-type, included as primary features of the megachiropteran 
("regime frugivore") type a conical cardiac region (cardiac vestibule of 
Grasse), an extensive pyloric tube, and a spacious fundic caecum. Addi- 
tionally, Grasse included an illustration (after E. Pernkopf) of the stomach 
of Pteropus vampyrus, an Old World frugivorous megachiropteran, that 
bears a resemblance to that of Sturnira and Artibeus in the features discussed 
above. Also, Grasse (1955) stated that the genus Nyctimene, a group of fig- 
eating megachiropterans of the subfamily Nyctimeninae, is exceptional 
among members of the suborder in possessing a simple globular micro- 
chiropteran-like stomach; Koopman (1967) and Walker (1968) noted that 
Nyctimene may include insects in the diet. 

Examination of more species and numbers of individuals of each kind 
will be required before generalizations regarding coevolution of gastric 
structure and food habits are truly conclusive. Further study may shed 
light on the possibility of parallel evolution in gastric morphology in the 
Megachiroptera and Microchiroptera, as suggested above. 

Subfamily Desmodontinae 
Desmodus rotundus (E. Georrroy-St. Hilaire, 1810) 
Specimens examined (3).— Three specimens from Nicaragua of 
dus rotundus minimis (KU 106260, 106262-63). 

Overall dimensions.— Greatest length 75.0-86.0; greatest breadt 



658 The University Science Bulletin 

gastroesophageal junction to pyloric sphincter 1.9-2.1; gastroesophageal junc- 
tion to apex of fundic caecum 72.5-83.0; breadth of pylorus at sphincter 
2.3-2.5. 

General morphology. — Stomach modified into narrow, elongate caecum 
of relatively uniform breadth originating at gastroesophageal junction, with 
diverticulation of gastrointestinal tract into stomach and duodenum (Fig. 
33); caecal pouch initially directed laterally (to the left), bending sharply 
caudad approximately 3 mm lateral to esophageal junction, continuing 
aborally to approximately midway along length, then recurving sharply 
craniad, and recurving again posteriorly just proximal to apex (Fig. 34); 
stomach uniformly narrow throughout most of length, dilated slightly 
near apex and slightly constricted within terminal 15 mm. Cardiac vesti- 
bule absent, squamous epithelium terminating symmetrically and abruptly 
at lesser curvature; aboral end of esophagus slightly expanded at junction; 
no evidence of anatomical cardiac sphincter; muscularis externa and muscu- 
laris mucosae of esophagus unmodified in dimensions or structure at junc- 
tion with lesser curvature. 

Pylo-duodenal junction and pyloric sphincter in close proximity and lo- 
cated at right angle to gastroesophageal junction; pylorus occupying region 
between pyloric sphincter and beginning of tubular fundic caecum (inclusive 
of mucous glands at gastroesophageal junction); valves of pyloric sphincter 
short and extremely thin, expanded slightly at apex, sphincter portion of 
muscularis externa composed only of circular muscle layer; remaining 
muscularis externa well developed (both layers); circular layer extremely 
complex and somewhat variable in depth; tunica propria extensively devel- 
oped throughout stomach as dense layer between muscularis mucosae and 
circular layer (heavy complements between loosely organized bundles of 
circular fibers, elastic fibers rare among collagenous fibers) ; stomach most 
heavily vascularized among species examined with extensive numbers of 
arterioles and venules within areolar connective tissue. 

Gastric mucosa unusual in that most glands reduced to simple alveolus 
or tubuloacinar type (Rouk, 1968); cellular complements reduced in number 
as compared to other bats examined. 

Musculature. — Wall relatively uniform in depth throughout caecum 
but slightly reduced in depth along lesser curvature in vicinity of gastro- 
esophageal junction. Longitudinal layer moderate in depth throughout 
gastric pouch and narrowest within dilated portion and at apex, occasionally 
observed broken into tightly packed clumps of fibers and sometimes ap- 
pearing as two layers. Circular layer always deeper than longitudinal layer 
at given point; fibers of circular layer organized throughout stomach 
(excepting incisura cardiaca) into normally large, loosely arranged, trans- 
versely-oriented bundles (Fig. 35) surrounded on upper surface and sides 



Studies of Stomachs of Selected American Bats 



659 




■^ 



Fig. 33. Diagram of the cardiac and pyloric junctions of the stomach of Desmodus 
rotitndtis. Dotted lines indicate the continuation of the fundic caecum; it was not possible to 
include it to scale in this figure; GEJ, gastroesophageal junction. For explanation of oth 
symbols, see Figure 3. 

Fig. 34. Semidiagrammatic representation of the stomach of Desmodus rotundas. 
the local dilation of the caecum near the apex. 



660 The University Science Bulletin 

by extensive amounts of connective tissue; circular muscle bundles slightly 
variable in depth, no bundling at incisura cardiaca and gastroesophageal 
junction. Muscularis mucosae in extremely thin sheets and no deeper within 
rugae than stomach wall proper. Lamina propria mucosae extensive superior 
to circular layer. Depth longitudinal layer 15 to 30, circular layer 30 to 100 
(usually 60 to 75), muscularis mucosae less than 10. 

Organization of gastric mucosa. — Rugal folds numerous in longitudinal 
parallel rows throughout entire length of stomach with infrequent trans- 
verse ridges connecting single pairs of rugae; folds in compressed zig-zag 
configuration throughout stomach; extensive variation in height of rugae 
but individual folds uniform in height throughout length. 

Zone of cardiac mucosa here considered absent as mucous glands at 
pyloric sphincter and gastroesophageal junction indistinguishable morpho- 
logically and histochemically; pyloric mucosa extending 3 to 4 mm laterally 
from pyloric sphincter to beginning of fundic tube (inclusive of mucosa at 
cardiac junction); transitional zone narrow, symmetrical, occupying proxi- 
mal-most 3.0 mm of fundic tube; remaining 95% of stomach lined with 
markedly reduced fundic mucosa unlike any other bat examined; gastric 
mucosa deepest within pyloric zone where glands truly tubular, decreasing 
progressively through pylo-fundic transition to shallow acinar glands of 
fundic tube. 

Cardiac glands. — Considered absent as mucous glands at gastroesopha- 
geal junction broad, relatively straight and tubular, indistinguishable from 
those at pyloric sphincter. 

Fundus. — -Fundic-type mucosa occupying nearly all of inner lining to 
apex of caecum; uniformly shallow throughout; expression '"tubuloacinar" 
as applied to fundic glands by Rouk (1968) appropriate; glands pouched, 
short and broad (Fig. 36) ; simple branched tubular glands characteristic of 
mammalian fundic mucosa not observed. Depth of fundic mucosa 60 to 70. 

Chief cells normally within basal third of gland, maximally to midpoint; 
cells pyramidal, narrow and elongate; nuclei small, rounded, and always 
located basally. Argentaffin cells present although extremely rare; more or 
less triangular, broader than those of other species examined; nucleus small 
and ovoid; peripheral granulation not observed. 

Parietal cells not abundant, sparsely distributed from base (occurring 
among chief cells) to gastric pit; uniformly large and usually circular; not 
abundant enough to allow examination of intraglandular size gradient 
although no small parietal cells observed; nuclei greater in diameter than 
those of adjacent mucous neck cells and uniformly circular. 

Mucous neck cells most abundant element; cells small, more or less 
cuboidal, with small antero-posteriorly compressed nuclei; cells extending 
throughout middle 30-50% of gland. Gastric pits necessarily short and 



Studies of Stomachs of Selected American Bats 



661 



T»V'.<V : • e : . . ***** .-. * 4^** . k mtk 




my*'. > ~tf 














" 





Fig. 35. Stomach wall of Desmodus rotundas (lower fundus) showing lamina propr 
anil circular muscle (CM); note the tubuloacinar-type mucosal glands, extensive 
propria, and bundling of circular fibers. 

Fig. 36. Tubuloacinar fundic glands of Desmodus rotundas. The longitudin.i 
stomach is left to right. 



662 The University Science Bulletin 











:■ ■ ■-■:■ :■:■. if- / :■ :■■■ ■ 






■ 
if 

■i 


-< 



20u 







#' 



37 

Fig. 37. Surface epithelium of fundic mucosa in the stomach of Desmodtts rotundtis. 

sometimes simple pore within surface epithelium; distinctive and rather 
broad lumen throughout length of gland. Surface epithelium extremely 
thick (Fig. 37) throughout all fundic mucosa; cells narrow and elongate, 
with small slightly ovoid nuclei within basal third of cell, extensive mucous 
border along lumenal surface. Lamina propria extensive beneath and 
between fundic glands, collagenous fibers infrequent, however. 

Pylo-fitndic transition zone. — Occupying narrow band within uppermost 
portion of tubular stomach immediately to left of gastroesophageal junction; 
acquisition of zymogenic cells abrupt at terminal margin; definition of 
initial onset of sparse parietal cells difficult, although cells appear immediately 
within glands and are notably reduced in depth (latter character signifying 
beginning of transitional zone in this bat) ; reduction in glandular depth at 
initial margin fairly abrupt, most of transitional area composed of tubulo- 
acinar elements; gastric pits also decreasing abruptly in depth; lamina pro- 
pria consistent with that of fundus; depth of transitional glands 70 to 150. 

Pyloric glands. — Originating at gastroesophageal junction and extending 
aborally to distal edge of pyloric sphincter and laterally to initiation of 
caecal tube; short, broad, tubular glands composed of single cell type; not 
notably coiled, broadly expanded at base, slightly expanded at upper limit. 
Cells narrow and elongate, numerous in cross-section of tubule; nuclei 



Studies of Stomachs of Selected American Bats 663 

small, spherical, located basally. Lumen extremely broad in base of gland, 
narrowing only slightly toward upper exit; glands deepest and narrowest 
at cardiac junction, shallowest at onset of oxynctic mucosa and at pyloric 
sphincter. Surrounding lamina propria and smooth muscle fibers con- 
siderably less abundant than within transitional or fundic mucosa. Depth 
of pyloric mucosa 150 to 210. 

Glands of Brunner. — Restricted to proximal 2.5 mm of duodenum and 
totally submucosal; distributional pattern similar to that in several other bats 
examined; pronounced external constriction at pyloric sphincter resulting 
from lateral expansion or bulge of glandular mass at proximal end of duo- 
denum; glands similar dimensionally and morphologically to those oi 
Artibeus; tubules broad, typically highly coiled; cells relatively numerous; 
nuclei slightly smaller and more nearly ovoid than those of Artibeus; 
lumen broad, increasing slightly in breadth toward crypts of Lieberkuhn; 
lamina propria mucosae and smooth muscle fibers of muscularis mucosae 
most extensive between tubules among bats examined. 

Remarks. — Several authors including Huxley (1865), Grasse (1955), 
Schultz (1965), and Rouk (1968) have previously illustrated and briefly 
discussed the topography and gross morphology of the stomach of Desmodus. 
Additionally, Schultz illustrated the gastroesophageal and pyloric areas of 
a closely related species of vampire, Diphylla ecaudata. Grossly, the stomach 
of Diphylla differs from that of Desmodus primarily in having a more 
laterally-displaced pyloric sphincter relative to the gastroesophageal junction 
(after Schultz), with a resulting probable increase in the distribution of 
pyloric mucosa. Robin (1881) noted that the caecum of Diphylla averages 
somewhat shorter than that of Desmodus. Of the previous investigators, 
only Rouk has endeavored to describe the composition of the gastric mucosa; 
my account differs somewhat in content from his. 

The elongate tubular stomach and the tubuloacinar-type mucosal glands 
are unique among bats examined. Because partly or wholly digested food 
materials that are present in the plasma of the ingested blood meal may 
represent the essential food source of vampire bats, it is not surprising to 
find a secondary reduction of mucosal cellular elements responsible for 
secretion of materials active in digestion of albumunous foods. Nutritional 
elements utilized by vampires and the mechanisms of intestinal absorption 
are little understood, thus retarding complete evaluation of this striking 
reduction of gastric mucosa. 

Pronounced segregation of circular muscle fibers into loosely arrange 
bundles, with prominent amounts of interposed loose connective I 
ably permits adequate stretching of the caecal wall to allow 
rather large blood meals (see Wimsatt and Guerriere, 1962). 
these authors noted an extensive condensation of the blood 



664 The University Science Bulletin 

stomach, with copious urine excretion following ingestion. The extensive 
vascularization of the stomach observed in Desmodtts accounts morphologi- 
cally for the rapid and distinctive fluid absorption from the stomach of this 
species. 

Family NATALIDAE 
Natalus stramineus Gray, 1838 

Specimens examined (1).— One specimen from Nicaragua of Natalus 
stramineus saturatus (KU 106280). 

Overall dimensions— Greatest length 7.3; greatest breadth 3.2; gastro- 
esophageal junction to pyloric sphincter 3.9; gastroesophageal junction to 
apex of fundic caecum 33\ breadth pylorus at sphincter 0.9. 

General morphology. — Stomach small, markedly comma-shaped; gen- 
erally similar to stomach of Plecotus in gross morphology, but strikingly 
like stomach of soricids described by Myrcha (1967); fundic caecum mod- 
erate in size, not directed anteriorly (but laterally), apex narrowly rounded 
with no partitioning of caecum by folds of musculature. Well-developed 
cardiac vestibule (Fig. 38) formed partially of dilated lower terminal end 
of esophagus (Fig. 39), hence uppermost portion of vestibule lined with 
stratified squamous epithelium (unique among bats examined); esophagus 
entering stomach directly anteriorly, hence no angle of entrance and no 
incisura cardiaca; cardiac vestibule grades gently into lesser curvature; 
junction with stratified squamous epithelium abrupt and inferior to dila- 
tion of terminal end of esophagus, cardiac orifice thus extremely broad; 
stratified squamous epithelium smooth at junction with gastric mucosa; 
no evidence of active cardiac sphincter. 

Tubular stomach distal to cardiac junction long and narrow; muscula- 
ture as well as pyloric sphincter better developed on greater than on lesser 
curvature; pronounced sulcus intermedius proximal to pyloric sphincter; 
stomach only slightly curved on the whole, tubular region relatively straight 
as compared to other bats examined and lying in plane nearly perpendicular 
to plane of esophagus; slight antero-lateral inflection of terminus originating 
at sulcus intermedius. Marked external constriction at pyloric sphincter; 
sphincter includes muscularis mucosae and circular layer; sphincter longer 
and narrower on greater curvature. Lamina propria well represented be- 
neath muscularis mucosae in rugae, otherwise scant, never abundant above 
muscularis mucosae layers. 

Mucosal folds well developed throughout stomach; musculature sub- 
stantially thicker in greater than lesser curvature. 

Musculature. — Stomach wall extremely thin throughout lesser curvature 
from apical arc of fundic caecum to just proximal to pyloric sphincter. 
Oblique muscle layer (muscularis mucosae) unusually thick and dense in 
midregion between gastroesophageal junction and pyloric sphincter on 



Studies of Stomachs of Selected American Bats 



665 




2mm 



Fig. 38. Mid-longitudinal representation of the stomach of Natalus stramineus. For ex- 
planation of symbols, see Figure 3. 



lesser curvature, at one point as thick as circular and longitudinal layer 
combined, formed of typically thin sheets throughout remainder of stomach. 
Sheets of muscularis mucosae within esophagus continuous with those of 
stomach with no alteration of morphology. Musculature (externa) of apex 
and lower surface of fundic caecum three times thickness of that of lesser 
curvature wall; longitudinal layer as deep as circular layer in much of 
fundic caecum. Circular muscle layer of greater curvature deep throughout, 
with some "bundling" in area adjacent to fundic caecum; longitudinal 
layer shallow; marked local thickening of circular muscle layer just proxi- 
mal to pyloric sphincter (only within greater curvature). Depth longitudi- 
nal layer 10 (lesser curvature) to 45, circular layer 30 to 140, muscularis mu- 
cosae 4 to 15 (lesser curvature). 

Organization of the gastric mucosa. — Mucosa gathered uniformly into 
parallel, longitudinal rows throughout all of stomach with no interdigita- 
tion of ridges; mucosa deepest along lesser curvature especially in mid- 
region at base of cardiac vestibule; only slightly reduced in depth at apex 
of fundic caecum and considerably less so than in Plecotus with no reduc- 
tion in frequency of rugae at apex; pyloric mucosa deep (deeper than most 
of fundic complement); fundic mucosa deep at junction with esophagus, 
uniform in depth throughout most of cardiac vestibule (in contrast to 
shallow, variable cardial mucosa of Plecotus); cardiac glands unusual 
both morphology and distribution (found beneath esophageal epi 
at cardiac junction as well as in stomach proper); fundic mucosa occup) 



666 



The University Science Bulletin 



, 1 m m , 




39 




* r 









l 



; 



'* 




« 






• - 
- 

. > - 



40 



Fig. 3 ( ). Photograph showing cardiac vestibule of Natalus stramineus. Note the narrow 
wall of the lesser curvature in contrast to the greater curvature. 

Fig. 40. Large cardiac glands at the gastroesophageal junction of Natalus stramineus. 



Studies of Stomachs of Selected American Bats 667 

most of internal surface area of stomach (as in other bats) but unusual in 
being most highly developed within fundic caecum; pylo-fundic transi- 
tion zone extensive as zymogenic cells limited in distribution; pyloric 
mucosa well represented and more extensive in distribution than most 
species examined. 

Glands of Brunner relatively sparse, present in small noncontinuous 
clumps in proximal 3 mm of duodenum and superior surface of pyloric 
sphincter. 

Cardiac glands. — Small band of highly convoluted tubular glands ad- 
jacent to, and beneath, lower limit of stratified squamous epithelium at 
gastroesophageal junction; mostly superior to muscularis mucosae but 
found also between the two layers of this tissue in uppermost area of 
distribution under esophageal lining. Glands extremely large (Fig. 40), 
two to three times breadth of adjacent gastric glands; lumen extremely 
broad, superior-most glands appear to empty through pits formed of modi- 
fied stratified squamous epithelium; glands composed of single cell type 
indistinguishable in morphology and staining reactivity (refer to section 
on histochemistry of cardiac glands) from cells of glands of Brunner; 
cardial cells large and with extensive mucous border; nuclei small, spherical, 
juxtaposed to basement membranes; cytoplasm agranular and stains ex- 
tremely pale pink with eosin Y; connective tissue not abundant but more 
so than that observed between glands of fundus; transitional glands be- 
tween cardiac and fundic mucosa difficult to identify, at best an extremely 
narrow zone characterized by occasional parietal cell in base of gland. 
Depth of cardiac glands 100. 

Fundus. — Extensive variation in depth and complements of cellular 
elements; true fundic glands best developed within fundic caecum and 
along lesser curvature near gastroesophageal junction; extensive variation 
in depth with deepest glands along (1) upper wall of fundic caecum, (2) 
within cardiac vestibule, and (3) at apexes of rugal folds; fundic mucosa 
slightly reduced in depth at extreme apex of fundic caecum with marked 
reduction in furrows between adjacent fundus 95 to 230. 

Zymogenic cells in highest concentration within, and in areas immedi- 
ately adjacent to fundic caecum; occupy lower 50% of gland in most 
areas, some areas of fundus with chief cells in lower 20% of gland (through- 
out much of greater curvature); chief cells gradually reduced in number 
lateral to plane of gastrosophageal junction in direction of pyloru 
pyramidal to trapezoidal or polyhedral, with large, basal, spherica 
basal region of gland including chief cells bulbous, parietal cells 
mon among chief cells. Argentaffin cells positively identified 
among basal-most cells; morphology essentially that as describee 



668 The University Science Bulletin 

field (1953) for argentaffin cells of human small intestine; cells long and 
narrow with nucleus displaced medially; dense peripheral granulation. 

Parietal cells abundant throughout most of fundus and transitional 
area; notably abundant in upper portion of glands resulting in extremely 
shallow gastric pits; cells circular to ovoid in lower areas of gland, becoming 
smaller and more rounded in upper portions of tubules; parietal cells re- 
duced substantially in frequency throughout apical arc of fundic caecum 
accounting for slight reduction in glandular height. 

Mucous neck cells typically small with unusually large nuclei; occur 
in small, tightly compressed clumps wedged among adjacent parietal cells; 
little cytoplasm. Gastric pits extremely shallow throughout all of fundus 
as well as transitional areas; pits rarely more than fifth total length of 
gland, even in fundic caecum; parietal cells occasionally found just be- 
neath surface epithelium suggesting essential absence of gastric pit and 
presence of simple "pore"; cells lining pits having typically large nuclei of 
lower mucous neck cells, nuclei becoming slightly laterally flattened toward 
surface of mucosa. All mucous cell components of glands of fundic caecum 
have more spherical nuclei than found in remainder of fundus. Surface 
epithelium thick; nuclei large, basal, club-shaped; cells in single layer. 
Connective tissue beneath glands nearly absent, glands often resting on 
mucularis mucosae. 

Pylo-fundic transition zone. — Region extensive owing to relatively lim- 
ited distribution of zymogenic cells as discussed previously; glands excep- 
tionally narrow and long, parietal cells abundant throughout most of length; 
parietal cells sometimes in continuous column within glands and in sub- 
stantially greater numbers than mucous neck cells; loss of zymogenic cells 
at beginning of zone gradual; loss of parietal cells at terminus of area also 
gradual with one or a few cells persisting in proximal-most pyloric glands; 
gastric pits as shallow as those of fundus proper; surface epithelium un- 
changed over that of fundus; little or no progressive decrease in depth 
except for slight decrease at point where parietal cells finally disappear; 
depth ISO to 220. 

Pyloric glands. — Zone of pyloric glands relatively extensive in comparison 
to total internal surface area of stomach, best developed on lesser curvature; 
glands generally narrow at origin of distribution, increasing in breadth toward 
termination at onset of glands of Brunner in duodenum. Glands composed 
of single cell type in single layer; nuclei large and spherical throughout. 
Gastric pits deep to base of gland (as characteristic of pyloric glands, mu- 
cous border of glands narrow resulting in broad glandular lumen; cells 
considerably larger than mucous neck cells of fundic glands; surface epi- 
thelium deeper than that of fundus owing to much larger epithelial cells; 
glands terminate at beginning of Brunner's glands on distal surface of 



Studies of Stomachs of Selected American Bats 669 

pyloric sphincter. Lamina propria more abundant at base of, and between, 
pyloric glands than in fundic mucosa. Depth of pyloric mucosa 200 to 230. 

Glands of Brunner. — Occur in discontinuous patches beneath muscularis 
mucosae in proximal 3 mm of duodenum; ring of glands occupy entire distal 
surface of pyloric sphincter, decreasing gradually in abundance to terminate 
0.5 mm from sphincter; other smaller patches of three to six glands occur 
distal to main group; proximal-most glands extremely large with spherical 
and peripheral nuclei and extensive cytoplasmic complement; more distal 
isolated clumps considerably smaller and nuclei always small and markedly 
flattened, lumen much narrower. Cells and glands indistinguishable from 
previously described "cardiac" glands. Brunner's glands clearly different in 
both distribution and morphology from other bats examined. 

Remarks. — The stomach of Natalus stramineus is unique among bats 
examined (and notably similar to stomachs of soricid insectivores studied 
by Myrcha, 1967) in possessing a terminal pyloric section oriented nearly 
directly laterally with little recurvature craniad. Also, N. stramineus pos- 
sesses the largest and most distinctive cardiac vestibule among carnivorous 
bats studied. 

The cardiac gland zone, although typically narrow, is distinctive in 
several respects. These glands are highly convoluted and extremely broad, 
a combination of features observed only in Natalus. Several tubules were 
observed to open through pores within the squamous epithelium at the 
gastroesophageal junction; thus, the cardiac glands lie partially beneath 
esophageal epithelium. Additionally, tubules of cardiac glands in cross- 
section are nearly indistinguishable from the glands of Brunner of the 
same bat. Bensley (1905) and others have noted unusual similarity between 
cardiac glands and Brunner's glands in several species of mammals, although 
the present case appears to be the first reported for bats. In all other species 
examined, the two glandular types were clearly distinctive in morphology 
within each bat. 

The fundic mucosa resembles most closely, among bats examined, that 
of Plecotus in distributions and morphology of cellular components, as well 
as in pattern of mucosal reduction in the fundic caecum. Natalus and 
Plecotus differ, however, in distribution of different mucosal elements; the 
pyloric zone is more extensive in Natalus, shifting the transitional area 
medially, thus reducing the extent of the fundic mucosa relative to that 
observed in Plecotus. Additional specimens of Natalus must be studied in 
order to verify the tentative conclusions presented here. 

Family VESPERTILIONIDAE 
Plecotus townsendu' Cooper, 1837 
Specimens examined (3).— Three specimens of Plecotus townsen 
lescens from South Dakota. 



670 



The University Science Bulletin 




2mm 



Fig. 41. Mid-longitudinal representation of the stomach of Plecotus townsendii. For ex- 
planation of symbols, see Figure 3. 



Overall dimensions. — Greatest length 8.9-9.5; greatest breadth 3.1-3.8 
(inclusive of cardiac vestibule); gastroesophageal junction to pyloric sphinc- 
ter 2.5-2.7; gastroesophageal junction to apex of fundic caecum 4.6-5.1; 
breadth pylorus at sphincter 2.0-2.1. 

General morphology. — Stomach saccular and reniform (when empty), 
notably symmetrical with cardiac vestibule at midpoint of lesser curvature 
(Fig. 41); caecum extensive but unpouched, occupying nearly half length 
of stomach; apex of caecum relatively blunt and directed antero-laterad, 
upper limit of caecum lying approximately in same plane as pyloric sphinc- 
ter and posterior limit of cardiac vestibule; caecum unusual in being long, 
rather pointed, directed somewhat craniad, and extending a distance equal 
to, or greater than, tubular stomach distal to cardiac junction; cardiac 
vestibule moderate in height, expanding gradually into lesser curvature; 
distance from esophageal junction to apex of fundic caecum greater than 
that to pyloric sphincter; junction having stratified squamous epithelium 
of esophagus abrupt, no suggestion of active muscular cardiac sphincter; 
incisura cardiaca absent (esophagus entering stomach perpendicularly). 
Terminal stomach distal to esophageal junction relatively short and broad; 
musculature of pyloric sphincter (Fig. 42) relatively uniform in mass 
throughout, but slightly more massive on greater than on lesser curvature, 
thicker along luminal edge than elsewhere; musculature of sphincter com- 
posed exclusively of inner circular layer plus thin sheets of muscularis 



Studies of Stomachs of Selected American Bats 671 

mucosae that line sphincter mass; notable external constriction circum- 
scribing pyloric sphincter; glands of Brunner moderate in abundance and 
restricted to distal surface of pyloric valve musculature and proximal-most 
2 mm of duodenum. Esophageal junction in a plane distinctly anterior to 
that of pyloric junction owing to moderately well-developed cardiac vesti- 
bule (Fig. 41); cardiac vestibule and pyloric orifices directed in approxi- 
mately same direction, terminal stomach at nearly right angle to plane 
of remaining stomach; stomach generalized in gross morphology with 
little apparent specialization of structure. 

Musculature. — Muscle layers generally moderate in depth when com- 
pared to other bats examined; total muscularis externa slightly thicker 
within lesser than greater curvature, circular layer notably thickened in 
terminal half of pyloric tube (especially thick in greater curvature) ; circular 
muscle layer always substantially thicker than longitudinal layer except 
in extreme apical tip of fundic caecum where layers approximately equal in 
depth; ganglia of Auerbach's plexus rare; extensive bundling of circular 
muscle occurs throughout proximal two-thirds of greater curvature; no 
suggestion of a muscular cardiac sphincter; depth longitudinal layer 15 to 
25, circular layer 40 (caecum) to 160 (pyloric tube), muscularis mucosae 
less than 10 to 15. 

Organization of gastric mucosa. — Mucosa clearly gathered longitudinally 
throughout sides of stomach (dorsal and ventral surfaces of organ in situ) ; 
rugal folds exceptionally numerous, long, narrow, and notably high; folds 
gathered uniformly longitudinally throughout all of terminal portion; folds 
clearly oriented transversely at apex of fundic caecum (Fig. 43) and within 
greater curvature up to onset of terminal stomach, and occurring as deep 
"pyramids" of mucosa lying directly along arc of greater curvature and 
wall of fundic caecum; folds of cardiac vestibule directed longitudinally; 
folds greatest in breadth in pyloric tube, decreasing gradually in breadth 
toward apex of caecum; folds of fairly uniform depth throughout sides, 
zig-zag orientation in lower half of caecum (Fig. 44), notably deeper along 
arc of greater curvature; slight increase in depth of folds in lesser curvature; 
substantial reduction in gastric mucosa only in apex of fundic caecum where 
folds are short, narrow, and relatively sparsely distributed; folds along arc 
of greater curvature in unusual configurations (extremely long, narrow 
"fingers" often expanded at their apexes) (Fig. 43). 

Well-defined cardiac, fundic, and pyloric glands present; cardiac zone 
extremely narrow (one or two glands in breadth); pyloric zone rel; 
narrow; fundo-pyloric transition area relatively broad as transition to pylon 
gland mucosa is gradual; fundic mucosa occurring throughout 
stomach and cardiac vestibule. 

Cardiac glands.— Junction with stratified squamous epithelium 



672 



The University Science Bulletin 



,0,5mm , 




• - 



'"f-7 








'43 



Fig. 42. Pylo-duodenal junction of Plecotus townsendii. 
Fig. 43. Mid-longitudinal section of stomach of Plecotus townsendii showing short, broad, 
rugal folds of greater curvature. 



Studies of Stomachs of Selected American Bats 673 

glands long and narrow, with extremely limited distribution, composed of 
single cell type; mucous cells of somewhat variable morphology throughout 
gland as basal cells of lower third of tubule small, compact, rectangular, 
more numerous than upper cells and with nuclei peripheral and spherical; 
cells of upper two-thirds with progressively more laterally flattened nuclei 
and cells more elongate than in basal elements; gastric pits always broad 
and increase in breadth toward luminal surface; glands increase slightly 
in depth from cardiac junction to transitional fundic zone. Transitional 
zone from cardiac to fundic mucosa narrow with infrequent parietal cells 
in lower half; gastric pits relatively deep until onset of true fundic glands. 
Depth of cardiac glands 200 to 300, breadth of cardiac zone 60 to 100. 

Fundus. — Glands relatively short and broad throughout most of fundus, 
long and narrow only on lesser curvature within (and lateral to) cardiac 
vestibule; fundic mucosa best developed on lesser curvature but notably 
reduced within extreme apex of fundic caecum; all cellular components 
large and relatively few in number. Depth of fundic glands 80 to 180. 

Basal zymogenic cells large, pyramidal with spherical nuclei; may 
occupy up to 60% length of gland but usually about 40%; cells rather 
uniform in size throughout stomach and reduced in size and number only 
in apex of elastic stomach; chief cells form basal "bulbous" expansion of 
gastric gland. 

Parietal cells relatively large and usually spherical (but occasionally 
ovoid) with large, spherical nuclei centrally located in cell; greatest density 
of cells in medial half of gland occurring extremely infrequently among 
chief cells and most often in clumps of three or four but occasionally as 
many as 20; marked increase in cell size from bottom to top of distribu- 
tion (uppermost cells extremely large), whereas nuclear size remains 
constant; only marked decrease in numbers in terminal half of fundic 
caecum (with corresponding increase in cell size), slight increase in rela- 
tive numbers in pylo-fundic transition area; intercellular caniculi pro- 
nounced. 

Mucous neck cells interspersed singly or in small clumps among parietal 
cells and most common just above chief cells (where parietal cells less 
numerous) and, of course, in necks of gastric pits; little if any size variation 
within gland; nuclei spherical, slightly smaller than those of parietal cells. 

Argentaffin cells extremely rare in base among chief cells; small and 
with peripherally located granulation but otherwise clear cytoplasm; not 
observed elsewhere in stomach. 

Gastric pits relatively deep; mucous pit cells large with spherical 
which become slightly flattened laterally in upper portion ol pii 
usually occupy 40-50% length of gland and occasionally three-foi 
length. Surface epithelium thick; nuclei extremely large and ovoid, often 
"club-shaped." 



674 



The University Science Bulletin 




l 



0,1mm 



m?'„ £ 




Fig. 44. A longitudinal section showing mucosal folds (rugae) in stomach of Plecotus 
townsendii. 

Fig. 45. Pyloric glands of Plecotus townsendii. 



Studies of Stomachs of Selected American Bats 675 

Pylo-fundic transition zone. — Transition to pyloric glands gradual and 
zone relatively broad; abrupt loss of zymogenic cells, whereas parietal cells 
reduced in numbers only gradually, those remaining in terminal portion 
of transition large; mucous cells increase in frequency as parietal cells 
reduced; no change in glandular depth over that of fundus; transition zone 
occupies approximately angular portion of stomach; depth 120 to 170. 

Pyloric glands. — Extremely broad and massive (Fig. 45), each gland 
expanded into giant bulb at base; lamina propria between adjacent glands 
substantially better developed than in fundic glands with smooth muscle 
fibers extending between glands within lower half of layer; extensive vari- 
ation in nuclear shape between glands, nuclei spherical, nuclear structure 
usually consistent within a given gland; single cell type present with 
exception of occasional argentaffin cell in extreme base; surface epithelium 
identical to that of fundus. Glands near pyloric sphincter with narrow, 
sharply-compressed cells resulting in extremely broad openings of gastric 
pits into lumen, hence surface epithelium reduced to narrow "clumps." 
Glands terminate abruptly at distal edge of pyloric sphincter; pyloric glands 
in relatively narrow zone substantially distal to pyloric bend. Depth of 
pyloric mucosa 100 to 230. 

Glands of Britnner. — Moderate in abundance and in single solid mass on 
entire distal surface of pyloric sphincter and proximal 1.5 mm of duodenum; 
highly convoluted, large with broad lumen; cells long, narrow; nuclei ex- 
tremely small, markedly flattened, and always positioned against basal 
surface (peripheral); cellular morphology uniform throughout gland; glands 
always submucosal and restricted to duodenum; muscle fibers of muscularis 
mucosae prominent between glands; open luminally between crypts of 
Liberkiihn. 

Remarks— The stomach of Plecotus townsendii is generalized in gross 
morphology, being symmetrical in shape, and having a simple fundic 
caecum and gradual curvature throughout. These features along with dis- 
tribution of zones of gastric mucosa are similar to those reported for several 
species of Insectivora by Myrcha (1967), particularly several species of Old 
World talpids and erinaceids. The pyloric and pylo-fundic transition zones 
are well defined and approximately equal in length. The contributing 
muscle layers in the stomach wall are relatively thick with increasing thick- 
ness in both upper and lower walls of the pyloric tube toward the pyloric 
sphincter. 

Rouk (196S) described the stomachs of Myotis velijer and Antrozous 
pallidas as being of the "simple type," and his figure of the stom 
Myotis velijer bears striking similarity to that of P. townst 
slightly in orientation of the cardiac vestibule and in distribution of p> 
fundic transitional mucosa. All three genera {Antrozous, Myoti. 



676 The University Science Bulletin 

have more or less symmetrical stomachs with a fundic caecum hooked 
craniad, smooth overall curvature, a gradual (rather smooth) pyloric bend, 
a moderately thick pyloric sphincter, and a short, narrow cardiac vestibule. 
A photograph published by Fischer (1909:124) of a "Situtspraparat der 
Baucheingegeweide von Plecotus aunt us" showed the stomach of this 
species to be virtually identical topographically to that described here for 
P. townsendii. 

Vizonyx vwesi (Menegaux, 1901) 

Specimens examined (1). — One specimen from Isla Partida, Gulf of 
California (KU 113998). 

Overall dimensions. — Greatest length 10.0; greatest breadth 7.1; gastro- 
esophageal junction to pyloric sphincter l.S; gastroesophageal junction to 
apex of fundic caecum 6.0; breadth pylorus at sphincter 2.6. 

General morphology. — Stomach large and generally tubular but ex- 
tremely rounded through fundic caecum and greater curvature (Fig. 46); 
expanded substantially in antero-posterior direction; dorsal surface convex; 
pyloric tube short. 

Cardiac vestibule short and, although cardiac glands comparatively 
abundant within vestibule, thick-walled and extremely narrow in relation 
to breadth of pylorus; vestibule increasing gradually and slightly in breadth 
from gastroesophageal junction to incorporation into lesser curvature, situ- 
ated approximately midway along lesser curvature; esophagus entering 
stomach at slight angle to lesser curvature, creating incisura between vesti- 
bule and caecum; incisura cardiaca short and broad with muscle layer 
prominently thickened locally and longitudinal layer unchanged; gastro- 
esophageal junction abrupt with all muscular complements continuous 
between esophagus and vestibule but with slight local thickening of muscu- 
laris mucosae and lamina propria mucosae at junction; no specific evidence 
of an anatomical sphincter. Gastric orifices in relatively close proximity to 
each other and generally equivalent in direction anteriorly, although 
vestibule slightly acute to pyloric tube. 

Fundic caecum extremely large and spacious but not truly saccular, 
gently rounded, nearly spherical; musculature reduced in depth within 
extreme apical portion of caecum only with no external or internal features 
suggesting anatomical demarcation from remaining stomach. 

Tubular portion at terminal point distal to plane of vestibule more than 
twice breadth of gastroesophageal junction, directed anteriorly to the right; 
pyloric curvature gradual and continuous with musculature prominently 
thickened throughout tubular portion (which tapers little to pyloric sphinc- 
ter). Musculature of pyloric sphincter short and massive, appearing more 
or less uniform in depth throughout; sphincter a simple, rounded projection 
of circular muscle from wall of pylorus, valve displaced more craniad 



Studies of Stomachs of Selected American- Bats 



677 




2mm 



Fig. 46. Mid-longitudinal representation of the stomach of Pizonyx vivesi. For explanation 
of symbols, see Figure 3. 



on greater than on lesser curvature; mucosa lining sphincter slightly re- 
duced in depth over that of remaining pylorus; deep but narrow external 
constriction at pyloric sphincter; glands of Brunner not abundant and 
restricted to duodenum; sulcus intermedius absent, sulcus angularis present 
but not pronounced. 

Pattern of depth of muscular wall similar to that observed in Plecotus; 
gastric mucosa deep throughout sides and apexes of rugal folds with pat- 
tern of local reductions of mucosa similar to that found in Plecotus; lamina 
propria mucosae considerably more prominent than in Plecotus, loose lamina 
with abundant collagenous fibers best developed throughout greater curva- 
ture and within rugae (including fundic caecum to apex), much less pro- 
nounced through lesser arc; elastic fibers generally infrequent (including 
fundic caecum) and lamina propria between mucosa and muscularis mu- 
cosae typically infrequent throughout. 

Musculature. — Longitudinal layer of uniform breadth throughoi 
of greater curvature and through fundic caecum to just lateral of : 
cardiaca, reduced to thin remnant sheet through incisura cardiaca, ca: 



678 The University Science Bulletin 

vestibule, and all of remaining lesser curvature to pyloric sphincter. Vari- 
ation in depth of circular layer within stomach basically similar to pattern 
in Plecottis; circular layer differing from that in Plecotits in being approxi- 
mately equal in depth within much of greater and lesser curvatures (con- 
sistently deeper in lesser as compared to greater curvature in Pelcotus) with 
proportionally greater thickening within pyloric tube, gathering of circular 
fibers into transverse bundles more frequent and bundles larger (although 
restricted to greater curvature and apex of fundic caecum as in Plecotits), 
and in having more extensive inter-bundle complements of lamina propria. 
Muscularis mucosae uniform in depth throughout stomach wall but notably 
deeper within rugae. Depth longitudinal layer 15 to 45, circular layer 35 
to 250, muscularis mucosae less than 10 to 25. 

Organization of gastric mucosae. — Pattern of gathering of mucosal folds 
notably similar to that of Plecotits; folds oriented strictly longitudinally 
throughout tubular portion, longitudinally directed through dorsal and 
ventral walls of fundic caecum with pronounced zig-zag orientation within 
lower one-half of caecum (as in Plecotits) ; short, transversely directed folds 
originating along greater curvature not nearly so abundant as in Plecotits, 
but transverse folds along lesser curvature of fundic caecum as in Plecotits. 
Rugae numerous, nowhere locally reduced in number; rugal folds com- 
paratively broader and deeper than those of Plecotus, probably as function 
of increased stomach size. 

Zones of gastric mucosa well defined with relatively few transitional 
elements from one major glandular type to another; region of cardiac 
glands extensive relative to all other bats examined, occupying major 
portion of short cardiac vestibule; fundic mucosa locally variable in mor- 
phology within extensive fundic stomach, lining all of remaining stomach 
except pyloric tube; transition from fundic to pyloric mucosa relatively 
abrupt, with zone narrow as compared to Plecotits and two other vesper- 
tilionids previously examined (Rouk, 1968), confined to narrow area within 
pyloric curvature; pyloric glands extremely large and prominent, occupying 
most of tubular stomach distal to pyloric bend. Mucosa consistently deep 
throughout stomach, reaching greatest depth within pyloric zone; pro- 
nounced reduction in depth of glands within furrows of fundic zone only 
(similar to condition of Plecotits), pattern of reduction of mucosal depth 
on sides and surfaces of rugae within apical arc of fundic caecum similar 
to pattern in Plecotits. 

Cardiac glands. — Glands occurring in broad, deep zone 1 mm in length 
at gastroesophageal junction (Fig. 46); glands extremely broad, being three 
times breadth of adjacent fundic glands; cells within basal fourth small, 
elongate and tightly packed within basal arc; nuclei usually peripheral, ex- 
tremely small, spherical, mucous border narrow; cells within upper three- 



Studies of Stomachs of Selected American Bats (>7 { ) 

fourths becoming progressively more elongate (increase in mucous border) 
toward surface epithelium (with nuclei unchanged in size although modi- 
fied to prominently flattened discs located in juxtaposition to basement 
membrane) ; lumen always extremely broad, especially within bulbous 
basal expansion as mucous border of basal cellular framework narrow; 
glands relatively uniform in breadth superior to basal portion; transitional 
zone to fundic mucosa a narow belt of two or three cardiac-type glands 
with infrequent basal parietal cells; areolar connective tissue extensive basal 
to (and between) glands with large numbers of inclusive collagenous 
fibers (restricted to cardiac zone) ; depth increasing slightly from gastro- 
esophageal junction to junction with fundic mucosa; depth of cardiac 
mucosa 250 to 300. 

Fundus. — Extensive variation in morphology and depth of glands; best 
developed in both depth and complements of cellular elements within lesser 
curvature; significantly reduced in depth between rugae of greater curva- 
ture but only slightly between rugae within all of fundic caecum; reduction 
in depth usually involving reduced numbers of cells, rarely reduction in 
dimensions of cellular components; glands usually long and narrow, rela- 
tively straight and tubular, distinctively expanded basally, thereafter uni- 
form in breadth to surface epithelium; gastric pits also highly variable in 
depth and breadth. Depth of fundic mucosa 150 to 280 (lesser curvature). 

Zymogenic cells occupying lower 40% of gland throughout lesser curva- 
ture (especially abundant within lesser curvature) and most of upper 
two-thirds of tubular portion of fundus; notably reduced in number 
throughout glands of lower wall of fundic caecum and midregion of greater 
curvature, increasing again in number in greater curvature within tubular 
portion; cells large with small, ovoid nuclei located variably within basal 
third of cell; areas in which zymogenic cells reduced usually having cells 
only within lower 10-15% of gland (argentaffin-type cells not observed 
but assumed present). 

Parietal cells most abundant in lesser curvature and cardiac vestibule; 
most common within upper 60% of gland, but scattered and less frequent 
among chief cells and moderately abundant through apical arc of fundic 
caecum (within glands on rugae) ; reduced in abundance and density on 
dorsal and ventral surfaces in upper portion of gland throughout 
body of stomach (excluding tubular portion); reduction corresponds 
1) increase in depth of gastric pits, 2) reduction of overall glandula 
and 3) marked drop in number of chief cells; marked lower to upper 
size gradient in parietal cells, ranging from small among basal ch 
to those nearly breadth of tubule just beneath necks of gastric pit: 
cells averaging large and spherical with relatively small, usual 
nuclei; intercellular caniculi not observed although assui 



580 The University Science Bulletin 

Mucous neck cells infrequent among closely packed and numerous 
parietal cells within upper portions of most fundic glands, but more nu- 
merous in midregion where parietal and chief cells less abundant; mucous 
neck cells identical in morphology to those of Plecotus except nuclei small 
(smaller than those of adjacent parietal cells). Gastric pits varying from 
10-25% length of gland throughout lesser curvature to up to 75% within 
glands of greater arc; increased depth of pits partially attributable to 
absolute increase in pit depth, as well as relative decrease in glandular depth 
as compared to lesser curvature; pit cells identical in structure to cells of 
cardiac and pyloric glands of this species, lumen broad. Surface epithelium 
approximately two-thirds depth of gastric pit epithelium, cells in tightly 
packed clumps at interim between adjacent gastric pits; clumps not ex- 
tensive in mass owing to great breadth of gastric pits; basal lamina propria 
sparse and with less abundant intertubule matrix than observed in either 
cardiac or pyloric mucosa. 

Pylo-fundic transition zone.— Narrow zone (Fig. 46) illustrating rela- 
tively abrupt transition to pyloric mucosa; reduction and subsequent loss 
of parietal cells fairly abrupt with parietal cells persisting in large numbers 
throughout nearly all of transitional area; loss of zymogenic cells even 
more abrupt (within breadth of two or three glands) at initial edge of 
transitional area; increase in depth of gastric pits gradual from about 30 to 
50% length of gland within transitional area, then abrupt increase to base 
at margin with pyloric mucosa; lamina propria mucosae prominent basally 
and between adjacent glands and confined in abundance to transitional belt, 
uncommon in adjacent fundic mucosa; depth 210 to 310. 

Pyloric glands. — Occurring within zone equivalent to that of Plecotus 
but notably different in morphology; glands usually large in both length 
and breadth (Fig. 47), similar to cardiac glands in dimension although 
slightly broader; cells within basal 10% similar to mucous neck cells, small 
and with spherical, centrally located nuclei and narrow inner mucous 
border; remaining 90% of length composed of giant, elongate, columnar 
cells with small, disc-like, crescent-shaped nuclei and a vast mucous border 
similar to upper cells of cardiac glands; glands distinctively bulbular at 
base, narrowed above basal portion, increasing in breadth only slightly to 
lumenal surface; lumen extremely broad throughout, broadest within basal 
region; pyloric mucosa restricted aborally to proximal surface of pyloric 
sphincter, not extending into duodenal lumen; lamina propria especially 
pronounced between glands, moderate basally. Pyloric, along with cardiac, 
glands representing most massive single mucosal-type glands observed in 
entire study. Depth of pyloric glands 250 to 340. 

Glands of Brunner. — Not abundantly distributed, occurring in narrow 
band within proximal 1.1 mm of duodenum and along distal surface of 



Studies of Stomachs of Selected American Bats 



68] 



"Sri 



V. 



< 




'W-«' 



X 



0.1mm 



V 



fckf* 




X 



V 



■ 












' / 



47 



Fig. 47. Pyloric glands of Pizonyx vivesi. 



sphincter; always submucosal and restricted to duodenum; typically con- 
voluted, uniform in breadth and relatively narrow; cross-sectional structure 
similar to that found in other bats examined, differing from Plecotus only 
in location and structure of nuclei; nuclei peripheral in cell (but not basal), 
compressed ovals (usually markedly compressed in Plecotus); tubules more 
heavily enclosed by surrounding smooth muscle fibers originating from 
muscularis mucosae than in other bats examined; exit from lumen narrow. 

Remarks. — The stomach of Pizonyx appears to be slightly altered from 
the basic insectivorous type. The general outline or configuration is similar 
to that of Plecotus and Rhynchonycteris, although the stomach oi Pizonyx is 
considerably more rounded and has a shorter and broader terminal portion 
than in either of the other two latter bats. The spaciousness of the fundic 
portion suggests a modification to assimilate large quantities of food at 
a single feeding. A noteworthy feature is the extensive distribution oi 
cardiac glands at the gastroesophageal junction. These glands are 
relatively straight tubular, and strikingly similar in morpholog; 
pyloric glands. 

The structure and local variation in depth of the tunica mus 
externa closely resembles that observed in Plecotus, although compl 



( ,S2 The University Science Bulletin 

of loose lamina propria within the muscularis externa and superior to the 
muscularis mucosae are substantially greater in Pizonyx. Although the 
valves of the pyloric sphincter of Plecotus are longer than those of the 
single specimen of Pizonyx examined, the asymmetrical orientation of the 
flaps is identical in the two species. 

Both the pylo-fundic transitional glands and the glands of Brunner are 
distinctively reduced in distribution as compared to Plecotus. Perhaps the 
narrow transitional zone is a consequence of advancement distally of basal 
zymogenic cells. 

The only feature in which Pizonyx clearly resembles Noctilio leporinus, 
another fish-catching bat (of which the stomach was observed macro- 
scopically only), is in the short, broad, gently arched terminal stomach, and 
this feature is shared also with Pteronotus parnellii. 

Family MOLOSSIDAE 
Molossus ater E. GeofTroy-St. Hilaire, 1805 

Specimens examined (2). — Two specimens from Nicaragua of Molossus 
ater nigricans (KU 106285, 106287). 

Overall dimensions. — Greatest length 12.1-13.7; greatest breadth 4.1-4.5; 
gastroesophageal junction to apex of fundic caecum 4.2-4.4; breadth of 
pylorus at sphincter 1.3-1.4. 

General morphology. — Stomach tubular with relatively short, narrowly 
rounded fundic caecum; generally slight anterior flexure of the stomach 
resulting in slight convexity of greater curvature; pyloric and cardiac orifices 
oriented so that pyloric sphincter substantially inferior to cardiac sphincter 
in stomach in situ; terminal portion tapers gradually to pyloric sphincter. 
Incisura cardiaca present but not well developed, esophagus entering stomach 
at angle; stratified squamous epithelium extending posteriorly to lesser 
curvature. 

Stomach extensive distal to cardiac junction (see Fig. 48) ; marked pyloric 
bend just proximal to pyloric sphincter, endpiece sharply recurved craniad; 
sulcus angularis present as constriction within greater curvature immedi- 
ately proximal to pyloric bend, corresponding constriction in lesser curva- 
ture less pronounced; pyloric constriction at sphincter pronounced and sym- 
metrical. Pyloric sphincter asymmetrical, mass of valve considerably larger 
on greater than on lesser curvature but sphincter not unusually large overall; 
valve funnel-shaped and directed into duodenum; contributing muscle in- 
cluding circular layer and muscularis mucosae. 

Musculature well developed throughout, except in fundic caecum; ex- 
tensive grouping oi fibers into bundles within circular layer; unusual and 
prominent "notch" in muscle wall of greater curvature; substantial reduc- 
tion in depth of wall in fundic caecum. Fundic caecum with apex directed 



Studies of Stomachs of Selected American Rats 



683 




2mm 



Fie. 48. Mid-longitudinal representation of the stomach of Molossus ater. For explanation 
of symbols, see Figure 3. 



slightly postero-laterad in contrast to the antero-laterad orientation in ves- 
pertilionids resulting in suggestion of general symmetry to stomach; caecum 
tapering uniformly laterally, never expanded on dorsal surface; nonmuscular 
submucosa extremely rare (with exception of prominent loose lamina 
propria overlying midventral "notch" within greater curvature), elastic 
fibers extremely sparse in most places and nearly absent within fundic 
caecum. 

Glands of Brunner well developed at, and directly proximal to pyloric 
sphincter, occurring in single mass at sphincter and as narrow layer within 
pyloric stomach; terminal end of bile duct dilated at junction with duo- 
denum, resulting in unusual ampule containing duodenal mucosa and 
submucosal glands of Brunner; stomach in general different in external 
morphology from those of other carnivorous species examined (see Fig. 48). 

Musculature. — Circular layer notably thickened at incisura cardiaca 
(thickened locally and confined to immediate region of gastroesophageal 
junction with muscle fibers gathered into long narrow sheets enclosed by 
septa radiating inwardly from depression at incisura); longitudinal layer noi 
correspondingly thickened at incisura; opposing musculature adjacent to 
incisura slightly thickened and gathered into bundles locally although ana- 
tomical cardiac sphincter probably not present. 

Circular layer generally thinner in lesser than in greater curva 
longitudinal layer always extremely thin, except in wall ol great* 
ture; circular layer irregular in depth, depth moderate at pyloric spf 
on lesser curvature, decreasing sharply at pyloric bend, narrow 



684 



The University Science Bulletin 



0.5 mm" 



WJt* 









?\A" 






.•»•■» 






#Sf Trt' > ***" - -ar yaw 













^t. 



*# * 






>• 




-^-**/. DP. 



A*J* 



■ft * >- 9 

f; ; IK*- -• it ^^ v " 50 



Fig. 49. Midventral cleft in greater curvature of stomach of Molossus ater. 
Fig. 50. Pyloric sphincter of Molossus ater showing Brunncr's gland cells. (BG) within the 
stomach. 



Studies of Stomachs of Selected American Bats 685 

lesser curvature (excepting cardiac junction) and a thin sheet with some 
bundling through fundic caecum, increasing to moderate depth through 
greater curvature to a point proximal to sphincter where layer again nar- 
rows; extensive bundling of circular layer occurring throughout most of 
greater curvature from apex of fundic caecum to midventral "notch." 
Large, distinctive depression in musculature of stomach wall of greater 
curvature in frontal half, marked by unusually well-developed submucos;' 
beneath muscularis mucosae and atypically deep fundic mucosa along fold; 
midventral notch clearly not an artifact as structural disfiguration of muscle 
layer present (Fig. 49). Musculature generally irregular in depth through- 
out stomach, especially through fundic caecum and greater curvature, par- 
tially owing to extensive gathering of circular fibers into irregularly shaped 
bundles. 

Muscularis mucosae typically double-layered, relatively well developed; 
uniform in depth throughout stomach. Ganglia of Auerbach's plexus 
abundant. Depth longitudinal layer 8 to 15, circular layer 15 to 250, muscu- 
laris mucosae 15. 

Organization of gastric mucosa. — Mucosa consistently gathered into 
straight, parallel, longitudinal rows throughout most of stomach, with 
slight zig-zag pattern in walls of fundic caecum lateral to plane of esopha- 
geal junction; only major transverse gathering is in greater wall of fundic 
caecum lateral to midventral notch. Extensive variation in depth of mu- 
cosa, generally deeper along lesser than along greater curvature and in 
tubular stomach (in contrast to caecum); also notable variation in depth 
within fundic mucosa which is deepest along lesser curvature and along 
musculature of midventral notch of greater curvature, but is shallow along 
apex of caecum with extensive variation in depth on both dorsal and ventral 
walls; marked reduction in number of folds occurring throughout fundic 
caecum as well; mucosa shallower in furrows than on sides ot rugae 
throughout stomach. 

Cardiac glands present and confined to typically narrow zone at gastro- 
esophageal junction; fundic mucosa extensive with marked variation in 
morphology and depth from one area to another, extending distally to 
midway between pyloric bend and cardiac junction on lesser curvature, 
somewhat further on greater arc; extremely narrow transition zone from 
cardia to fundus; pylo-fundic transition glands (Fig. 48) occupy area 
between true fundus and pyloric bend, although somewhat deeper 
pyloric tube on greater than on lesser curvature; transition zone approxi 
mately equal in total length on greater and lesser surfaces; pyloric gl 
in well-defined region and with unusual basal cell component 
identical to glands of Brunner; pyloric glands not extending di 
yond proximal surface of pyloric sphincter. 



686 



The University Science Bulletin 



■ *■*"* . i ^^ 







< >" 



















j 






>*" Ss/:-i&* ft t 1 



iff n sR 



* 



ii 
0.1mm 



*VM 






•, 




I 









Fig. 51. Glands oi Brunner (BG) within the duodenum and pylorus of Molossus ater 
(Periodic Acid SehifT Leucofuchsite Reaction). 

Fig. 52. Transitional zone of Molossus ater. Note the infrequently scattered parietal cells 

(PC). 



Studies of Stomachs of Selected American Bats 687 

Glands of Brunner well developed as submucosal glands on distal sur- 
face of pyloric sphincter and proximal 2 mm of duodenum (Figs. 50-51), 
occurring also in pyloric glands as mucosal elements, indistinguishable in 
morphology from Brunner's glands within duodenum (Fig. 50), also occu- 
pying dilated chamber at entrance of bile duct into duodenum; mucosa 
generally shallow throughout stomach. 

Fundic glands deepest in midregion between fundic caecum and tran- 
sition zone; cardiac glands longer than fundic glands; fundic glands 
abruptly decreasing in depth in comparison to adjacent cardiac glands; 
pyloric mucosal membrane slightly deeper than most of fundus, perhaps 
owing to underlying cells of Brunner's glands, and relatively uniform in 
depth (deepest at bases of rugae and between adjacent folds). 

Cardiac glands. — Glands contained within extremely narrow zone (100 
to 200 wide) at gastroesophageal junction; composed of single cell type, 
cells similar to those of surface epithelium of remaining stomach; glands 
simple, tubular, of branched type; cells differing from those of fundus in 
having consistently spherical nuclei throughout; cells small, long and 
narrow, extremely compact within gland with small, spherical nuclei; 
glands easily distinguishable from pyloric glands of this species on account 
of nuclear morphology; glands longer than fundic and pyloric glands and 
with extremely broad lumen; basal lamina propria scanty; depth of cardiac 
glands 300 to 360. 

Fundus. — Fundic region occupying all of saccular stomach and tubular 
area up to approximately midway between cardiac junction and pyloric 
bend. General pattern in distribution of fully developed fundic glands 
that of a parallelogram with upper facies directed toward fundic caecum; 
glands with well-developed chief cells extending further into fundic caecum 
along lesser than along greater curvature (possibly a developmental phe- 
nomenon due to flexion of the pyloric tube); glands relatively uniform in 
cell distribution throughout fundic zone, except glands in apex of fundic 
caecum which show marked structural reductions; fundic glands usually 
moderate in length and narrow, extensively convoluted; glands expanded 
greatly at lower end, less so in upper portion. Most fundic glands uni- 
formly and consistently devisible into three zones, defined by types of cells 
present; upper third composed of columnar epithelium lining deep gastric 
pits, middle third composed of parietal cells and mucous neck cells, lower 
third composed of chief and argentaffin cells; glands of middle and uppe 
fundic caecum marked by substantial increase in length and breadt 
gastric pits to 75% length of gland with lower fourth occupied by 
infrequent parietal cells mixed with mucous neck cells and occa 
chief cells at base; gradual reduction in number of chief ce 
gradual reduction in number of chief cells from midregion of 
apex of fundic caecum. Depth of fundic mucosa 140 to 260. 



688 The University Science Bulletin 

Zymogenic cells abundant basally in most of noncaecal fundic mucosa; 
large, pyramidal with large and spherical nuclei; cells always greater in 
length than breadth; argentaffin cells commonly observed among chief cells 
at base and narrowly triangular with extremely large, spherical nuclei, and 
peripheral argentafile granules. 

Parietal cells large and almost always circular with moderately sized 
spherical nuclei always centrally located; cells usually restricted in distri- 
bution to medial third and never abundant; size of cells uniform through- 
out distribution in given gland with little size increase from bottom to top 
of zone; found in basal fourth of tubules in fundic caecum but rarely ever 
observed among zymogenic cells. 

Mucous neck cells infrequent beneath base of gastric pit and scattered 
among parietal cells; nuclei larger than those of parietal cells, cytoplasm 
minute, cells irregularly shaped. Gastric pits deep; cells lining pits large, 
with nuclei becoming increasingly more laterally flattened toward surface 
of mucosa; surface epithelium extremely thick, cells large with extremely 
large nuclei that are only slightly compressed. Lamina propria always 
sparse beneath glands but somewhat more extensive beneath muscularis 
mucosae along greater curvature, owing to buildup of connective tissue 
around bundles of circular muscle fibers. 

Pylo-fundic transition zone (Fig. 52). — Broad zone occupying approxi- 
mately same area on both curvatures, although distributions not exactly 
comparable; loss of zymogenic cells gradual within fundus, zone of gastric 
mucosa lacking chief cells and maintaining parietal cells extending slightly 
distal to pyloric bend; loss of parietal cells in terminal area also gradual, 
but more abrupt than termination of chief cells proximally; parietal and 
mucous neck cells extending to base of glands, although parietal cells 
less abundant in base than in middle third; gastric pits, surface epithelium, 
and density of lamina propria equivalent to conditions observed within 
fundic mucosa; depth 220. 

Pyloric glands. — Glands occupy zone within most of stomach between 
pyloric bend and pyloric sphincter; deepening of gastric pits gradual through 
terminal portion of transition zone with parietal cells disappearing gradu- 
ally; basic pyloric gland morphology similar to that observed in other bats, 
but unusual in presence of apparent layer of glands of Brunner beneath 
entire distribution of pyloric mucosa and hence contributing to the depth 
of the mucosal membrane (see section on glands of Brunner) ; mucous 
cells of pyloric glands large; nuclei extremely large, becoming progressively 
more flattened laterally from bottom to top of gland; pyloric glands in- 
creasing in breadth from base of gland as cells increase in size, lumen also 
becoming progressively broader upwards; depth of pyloric glands irregular 
but generally deeper on lesser than on greater curvature; surface epithelium 



Studies of Stomachs of Selected American Bats 689 

extremely thick, slightly thicker than that of fundus; depth of pyloric glands 
120 to 200. 

Glands of Brunner. — Glands well represented as a single mass in sub- 
mucosal distribution along distal surface of pyloric sphincter and proximal 
2 mm of duodenum; also found in terminal dilation of bile duct (also 
submucosal) and directly beneath entire complement of pyloric mucosa 
(usually above muscularis mucosae) as thin layer of glands reduced in size; 
duodenal glands long and narrow, highly convoluted with small, pyramidal 
cells; nuclei large, peripheral, ovoid to substantially flattened laterally; cyto- 
plasm pale pink and diffusely granulated when stained with Harris' 
haemotoxylin and eosin; glands always submucosal; those within mucosa of 
pylorus approximately half diameter of duodenal complement (Fig. 51), 
highly convoluted, both mucosal and submucosal (observed between two 
layers of muscularis mucosae) although usually within mucosa; occurring 
also at base of mucosal membrane between adjacent pyloric glands; cellular 
morphology of pyloric complement identical to that in duodenum, lumen 
narrow in both. 

Remarks. — The tube-like stomach of Molossits presents a perplexing 
array of features; most features represent conditions observed in other 
insect-feeding bats. The stomach, however, although tubular, is less sym- 
metrical in outline than is that of other carnivorous bats and has the elon- 
gate terminal stomach usually observed in bats that are primarily or entirely 
herbivorous. The pyloric sphincter is moderately developed, asymmetrical, 
as in Plecotus. 

The gastric mucosa is uniformly gathered longitudinally through the 
simple, unpartitioned fundic caceum. The patterns of distribution of cellular 
elements within the fundic mucosa and variations within the fundic caecum 
are essentially as observed in Plecotus. The most noteworthy feature of 
the gastric mucosa is an extremely thin layer of narrow, convoluted, tubular 
glands which are histologically indistinguishable from the glands of Brunner 
in this species and which lie immediately beneath the zone of pyloric mu- 
cosa. These serous tubular elements are less conspicuous and presumably 
less abundant than the corresponding layer of glands observed in Sturnira. 
In one specimen of Molossits, the sub-pyloric glands appeared to be in dis- 
continuous masses near the proximal surface of the pyloric sphincter, sug- 
gesting less complete invasion into the stomach than observed in specimens 
of Sturnira. 

One unusual feature observed only in Molossits is a deep and promi 
transverse cleft in the stomach wall that is located along the greater cu 
ture. This depression is clearly not an artifact of fixation or 
or a contractional wave within the stomach musculature, because 
is accompanied internally by a distinctive break in the muscle 



690 The University Science Bulletin 

heavy local concentration of lamina propria, and slightly deeper gastric 
mucosa than in adjacent areas. It is further of interest that Rouk (1968), 
in his description of the stomach of a molossid, Tadarida brasiliensis, de- 
scribed and illustrated a single depression midway along the greater curva- 
ture that appears to be identical to that described here for Molossus. 

Rouk's description of the stomach of T. brasiliensis suggests further 
noteworthy similarities between that species and Molossus ater. The gen- 
eral stomach outline is similar, although curvature of the stomach anteriorly 
is greater in T. brasiliensis. The fundic caecum in both species tapers to a 
narrow apex, and the orientation and position of the gastroesophageal 
junction and structure of the pyloric sphincter appear to be identical in 
the two. Neither species has a cardiac vestibule, and both have an elongate, 
cranially reflected, terminal pyloric endpiece. 

HISTOCHEMISTRY OF THE GASTRIC MUCOSA 
AND DUODENAL JUNCTION 

The results that follow are limited to a survey of basic types of reactions 
for mucosubstances, and are to be regarded as necessarily preliminary to 
more detailed studies. To my knowledge, the present study (excepting 
the brief PAS-Alcian blue account of surface epithelial mucus of Rouk, 
1968) is the first discussion of comparative mucopolysaccharide histochem- 
istry of the gastric mucosa of bats. The following accounts are directed 
to the identification and discussion of neutral and acid mucopolysaccharides 
as demonstrated by toluidine blue O metachromasia, PAS, and Alcian blue 
8GX 300. No attempt was made to identify specific chemical end-groups, 
because the objective of the study was to examine and characterize mucin 
specificity in species of bats relative to each other. More detailed discussion 
of specific categories of mucin are to be found in the accounts of techniques 
to follow, where a particular technique was most appropriate in mucin iden- 
tification. 

Periodic Acid Schiff Reaction 

The PAS reaction is generally thought to indicate the presence of 
neutral mucopolysaccharides, and, because most mucins are apparently 
colored with this procedure, it also provides a general indication of overall 
mucous activity. Mucins stain red or magenta with this procedure. 

In any test for the presence of complex carbohydrate material, it is 
necessary to account for the potential reactivity of lipids and simpler carbo- 
hydrates. As a check for the presence of distinctive concentrations of gly- 
cogen, sections of stomachs of each species were subjected to digestion by 
salivary amylase for 10 minutes at room temperature, prior to oxidation 
by periodic acid. In no case was glycogen, or other material presumably 



Studies of Stomachs of Selected American Bats 691 

removed by exposure to amylase, found to be a contributing factor to the 
intensity of PAS reactivity in sites of production of mucin material. 

Using PAS, staining cells of the glands of Brunner always was uniform 
throughout the cytoplasm, with no apparent local areas of especially high 
intensity. Staining of surface epithelium was restricted to a narrow border 
of mucous droplets along the luminal surface (surface opposite to base), 
except for slight additional staining of fine cytoplasmic granules just be- 
neath the luminal mucous material in one species (Plecotus). Parietal cells 
stained uniformly throughout the cytoplasm, but mucous neck cells stained 
only within the cytoplasmic material between the nucleus and lumen of 
the tubule (outer portion of cytoplasm). All cells within the cardiac and 
pyloric glands, irrespective of location, stained only with the luminal portion 
of the cytoplasm. The above statements apply equally well to those cells 
that stained with Alcian blue, except for an apparently more complete 
staining of the mucous neck cell cytoplasm with Alcian blue in the case 
of Phyllostomits. 

A uniformly strong reaction characterizes the surface epithelial mucus of 
all species of bats with the exception of Desmodus, in which the reaction 
is only slightly reduced in intensity in comparison to the other species 
examined. An equivalent statement may be made concerning the gastric 
foveolae of the fundic, pyloric, and cardiac glands of Desmodus, excluding 
the lower portions of cardiac and pyloric tubules to be discussed later. 
Staining was noticeably stronger in the foveolae within the fundic caecum 
in contrast to the remainder of the fundus in two species, Glossophaga and 
Carollia. Duodenal goblet cells of all species examined reacted with a uni- 
formly high degree of intensity. 

The glands of Brunner reacted relatively weakly in Rhynchonycteris, 
Plecotus and Pizonyx, moderately in Artibeus and Sturnira, and strongly to 
extremely strongly in the remaining species. Identical qualities of reactivity 
in the glands of Brunner of the duodenum and the basal cell complexes 
within the pyloric stomachs of Sturnira (Fig. 28) and Molossus (Fig. 51) 
help to confirm their previous identification, on morphological grounds, as 
glands of Brunner. Genera of the family Phyllostomatidae (excluding 
Sturnira and Artibeus) and the single noctilionid consistently revealed the 
greatest degree of reactivity in the glands of Brunner. 

Deeper cellular elements of the fundic mucosa revealed highly varied 
reactivity to PAS. Mucous neck cells among parietal cells react consistently 
less than surface or foveolar epithelial cells, except for equivalent i 
in the case of Noctilio. The frugivores (Carollia, Sturnira, and Artibeu 
as a group, showed less reactivity within the mucous neck cells 
parison to carnivorous species, except for extremel; 
Desmodus and virtually no positive coloring at all in Plecotus i 



692 The University Science Bulletin 

Little mention has been made in the literature of PAS reactivity, positive 
or negative, of gastric parietal cells. Leblond (1950) noted that the parietal 
cells of albino rats stain "weakly" with a periodic acid-fuchsin-sulphurous 
acid procedure, approximately equivalent to PAS, and suggested that 
some mucin activity might be attributable to parietal cells. A clearly positive 
staining in parietal cells, over and above that of several other gastric elements 
that are clearly PAS positive, was observed in several species of bats. Only 
a trace of reaction was observed in carnivorous bats, except for a weak 
reaction in the blood-feeding Desmodus, and moderate staining in Molossits. 
The presumably partially nectivorous and partially carnivorous Glossophaga 
revealed moderately stained parietal cells, as did three frugivorous species. 

Although Lillie (1965) recorded positive PAS reactivity in the zymogenic 
granules within the salivary glands of man and several species of rodents, 
and in the pancreatic zymogen granules of man, rabbits, and some mice 
(but not in rats or guinea pigs), little or no information appears to be 
available regarding such reactivity in the zymogenic granules of gastric 
chief cells. Little or no staining was observed in gastric zymogenic cells of 
the bats examined, with the exception of a consistently strong (+4) staining 
of the zymogenic granules in two specimens of Noctilio. The significance 
of this result is unknown. 

As noted previously, the major portions of the cardiac and pyloric 
foveolae stain intensely with PAS, with little variability among species. 
In contrast, the lower one-third to one-tenth, depending upon the species, 
of the cardiac and pyloric tubules were found to always stain less intensely 
than the remainder of the tubule (excepting the equivalent reaction in the 
pyloric glands of Molossits). A reasonable explanation for this might be 
that less total mucous material is present in the basal portions of the 
tubules than within the upper border. However, little physiological evi- 
dence is available to confirm such a statement, and results of staining with 
Alcian blue, discussed beyond, tend to refute this idea as well. Loss of 
staining capacity within the basal portion is abrupt from one cell to the 
next, suggesting a specific loss of some cellular material within the basal 
area. The degree to which the basal area stains in contrast to the remainder 
of the tubule depends upon the glandular area under consideration, and 
the species examined. 

In the case of the cardiac glands, basal reactivity is only slightly less 
(+3 or +4) than in the remainder of the gastric tubule (+5), except for 
marked decreases (+1 or trace) in the cases of Plecotus and Pizonyx. A 
similar pattern was observed in the pyloric glands, but several additional 
species, all carnivorous, revealed marked reduction in basal reactivity. Basal 
portions of the pyloric glands of Plecotus and Pizonyx revealed little or no 
reactivity. Also, the pyloric glands of Rhynchonycteris and Molossits had 



Studies of Stomachs of Selected American Bats 693 

little reactivity in the basal area. All the phyllostomatids and Natalus re- 
vealed only slightly less basal staining in the pyloric glands in comparison 
with the remainder of the tubule. 

Overall production of neutral mucins, as revealed by PAS, appears to 
be relatively uniform among bats examined. There is clearly generalized 
reduction in intensity of PAS reactivity throughout the stomach and upper 
duodenum of Desmodus, although relatively slight in comparison to other 
bats. Species exhibiting especially high levels of secretion of gastric mucins 
demonstrated by PAS include Noctilio, Pizonyx, Molossus, and particularly 
by Phyllostomus. 

Because results of PAS are uniform among species for a number of 
glandular and cellular elements, it is difficult to draw conclusions regarding 
relationships between and among the bats studied. It is of interest to note, 
however, that certain pairs or groups of bats are nearly identical, yet unique, 
in reactivity in the spectrum and combinations of glandular or cellular 
types showing particular degrees of staining capacity. Pkcotus and Pizonyx 
are nearly identical in reactivity, differing slightly only in the basal portions 
of the cardiac glands. Sturnira and Artibeus are equally similar, as are 
Pteronotus and Phyllostomus. Members of the Phyllostomatidae are not 
equivalent in reactivity of all gastric regions examined. Carnivorous and 
partially carnivorous phyllostomatids can be separated from frugivorous 
kinds by the intensive staining in the lower mucous neck cells and weaker 
staining in the parietal cells. Glossophaga, principally a nectivore that also 
eats some insects and fruit, evidences staining of the lower mucous neck 
cells and nearly equally intensive staining of parietal cells. The most un- 
usual bats in terms of patterns of staining are Noctilio with heavily stained 
zymogenic granules, and Desmodus, which exhibits overall reduction in 
presence of PAS positive mucus. 

Reaction to Toluidine Blue O 

Toluidine blue O stains mucous material either orthochromatically 
(blue) or metachromatically (reddish-purple). Toluidine blue metachro- 
masia is thought to indicate the occurrence of sulfated acid mucopoly- 
saccharides. 

To test for the presence of hyaluronic acid which may stain metachro- 
matically, tissue sections were exposed to an aqueous medium of testiculai 
hyaluronidase, which presumably digests both hyaluronic acid and sev 
chrondroitin sulfates. No demonstrable alteration of local metachrom 
was observed in any species after 18 hours digestion at 40°C. The 
suggest absence of hyaluronic acid and chrondroitin sulfates 
distinctive components of certain metachromatically stainin 
study. 

A summary of the results of toluidine blue staining is show 



694 



The University Science Bulletin 



Table 1. Toluidine blue metachromasia. 



V 

a > 

Species O <-£ 

Rhynchonycteris naso + (tr) 

Noctilio labialis — 

Pteronotus parnellii — 

PhyUostomus discolor + 

Glossophaga soricina + 

Carollia perspicillata — 

Sturnira lilium — 

Artibeus lituratus 

Desmodus rotundus — 

Natalus stramineus — 

Plecottts townsendii + (tr) 

Pizonyx vivesi + 

Molosst/s ater — 

Control (Peromyscus) .... — 



Area of Reactivity 



c c 

■ — ! u 

O PQ 



-f-(tr) 

+ (tr) 
+ (tr) 
+ (tr) 

+ 



c c 
cC u 



ca v. 



S^ 



+ 
+ 
+ 



+ (tr) 


— 


-f(tr) 


+ 


+ 


+ 


— 


+ 


+ 


+ 


— 


+ (tr) 


+ 


— 


+ 



o — 

O 8 



+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 
+ 



Some slight variation in staining was noted in the glands of Brunner be- 
tween individual specimens of two species (Plecotus townsendii and Carollia 
perspicillata), but the reaction to this stain of specimens of all other species 
was uniform for a given gland. This observation complements the negative 
results of sulfate autoradiography of Jennings and Florey (1956) in surface 
epithelia of several species of rodents. Decidedly positive reactions were 
widespread in the foveolae of fundic glands only in Rhynchonycteris and 
PhyUostomus, and in the fundic caeca of Plecotus and Pizonyx, with slight 
reactivity in the fundic caecum of Glossophaga. Sections of the stomachs 
of two rodents used as controls {Peromyscus leucopus and laboratory-raised 
Mus musculus) yielded no metachromatic staining in either surface epithelia 
or gastric foveolae. 

The most consistently positive metachromatic staining was observed in 
the goblet cells of the upper duodenum. All species examined yielded a 
distinctly positive metachromasia. 

Well-defined patterns of positive metachromasia within systematically 
related groups or among bats with similar food preferences were not ob- 
served. It is of interest to note, however, that only PhyUostomus yielded 
a positive metachromasia in the deeper mucous neck cells. Also, the two 
specimens of P. discolor examined demonstrated the greatest overall meta- 
chromasia within the seven regions of the stomach considered, apparently 
indicating atypically well-developed sulfate mucous material (s) in that 
species. The results also suggest that noncarnivorous species, especially 



Studies of Stomachs of Selected American- Bats 695 

frugivores like Sturnira and Artibeus, contain comparatively less meta- 
chromatic material, and hence mucous material with sulfate groups, than 
do most carnivorous species. There are exceptions, however, and additional 
studies involving more species and numbers of species are required to test 
this point. Desmodus shows definite overall lack of gastric metachromasia, 
although perhaps for different reasons than the frugivores. Little is known 
regarding the physiological role of gastric mucins in general, and less with 
specific regard to those with sulfate ester linkages (if indeed toluidine blue 
metachromasia demonstrates such substances). 

All stomachs examined were devoid of gastric contents and, therefore, 
supposedly were in relatively equivalent physiological states with respect 
to mucus production; the variability observed in response of mucins to 
toluidine blue O probably reflects true differences among the species studied. 

Reaction to Alcian Blue 

The Alcian group of dyes, including Alcian blue, green, and yellow are 
thought to denote the occurrence of acid mucopolysaccharides by way of 
coloration of carboxyl groups. Mowry (1963) noted that Alcian blue closely 
approximated the results of Hale's reaction (Hale, 1946) for acid polysac- 
charides. 

The Hale colloidal iron procedure was tried twice on sections of stomachs 
of bats, but results were entirely negative for every species, perhaps a result 
of inappropriate fixation. 

Two different, although similar, Alcian blue tests were conducted. In 
the bats studied, reactions to the two procedures utilizing Alcian blue were 
generally comparable excepting where cardiac and pyloric glands were in- 
volved. 

Exposure to periodic acid prior to staining with Alcian blue was reported 
by Mowry (1963) to demonstrate the presence of glycogen. No distinctive 
concentrations of glycogen were found in my study by the use of this 
procedure. 

Goblet cells of the upper duodenum reacted strongly to both Alcian blue 
procedures in every species examined. Spicer (1960) noted that acid dif- 
ferentiation appeared to increase staining of sulfated mucins by Alcian blue; 
however, in light of the clearly positive reactivity in the absence of periodic 
acid oxidation, it appears safe to suggest that both sulfated and non- 
mucosubstances occur in the goblet cells of most, if not all, bats examined. 
These results are entirely consistent with those obtained by Spicer 
for three species of rodents and a rabbit. 

The duodenal glands of Brunner revealed strongly positive 
with Alcian blue in only one species, Carollia, in contrast to cor 
more widespread toluidine blue metachromasia in these glands ; 
species of bats (see Table 1). Staining of mucins wit 



696 The University Science Bulletin 

Brunner not only occurred in two specimens of Carollia but was exceedingly 
strong. Although positive reactivity might have been the result of sulfated 
mucosubstances known to be present by the results of toluidine blue tests, 
this prospect seems unlikely in the absence of positive results in any other 
species. A trace of Alcian blue staining was observed in Brunner's gland 
cells of Sturnira and Artibeus, but it is difficult to justify this as resulting 
from the presence of mucosubstances because there was no contrast with the 
surrounding tissue. The basal-most cell complex within the pyloric stomach 
of Sturnira revealed but a trace of staining, and that of Molossus revealed 
none, perhaps further justifying the identification of glands of Brunner 
within the stomachs of these two bats. 

A weak to strong reaction was recorded in the gastric surface epithelium 
of the cardiac stomach of all species examined excepting Sturnira, Artibeus, 
and Desmodus. The epithelium of Sturnira and Artibeus stained only with 
a trace, that of Desmodus not at all. Results were nearly identical between 
the two Alcian blue procedures, with no distinctive differences requiring 
evaluation. 

Surface epithelial cells of the fundus reacted positively to Alcian blue, 
although consistently slightly less than the cardiac epithelium in most 
species. Rhynchonycteris and Phyllostomus would seem to be exceptions 
to this trend by showing equality of reactivity in the two glandular regions, 
and Desmodus revealed moderate staining in the fundic epithelium in con- 
trast to none at the surface of the cardia. Several species demonstrated 
decreasing reactivity of the surface epithelium with Alcian blue through the 
terminal portion of the fundus and through the pyloric zone to the pyloric 
sphincter. These included all carnivorous species and Glossophaga, but 
not the three frugivorous species or the sangivorous Desmodus. This reduc- 
tion in staining quality probably represents a progressive decrease in reactive 
material from the midstomach toward the pyloric sphincter. This condition 
is especially interesting in light of the often observed increase in depth 
of the surface epithelium from the fundic mucosa through the pylorus. 

Reactions of cells within the cardiac glands are highly variable between 
portions of a single tubule, among species, and between the two procedures 
for certain species. As in the case of PAS, the upper neck as well as various 
relative percentages of the basal portion of tubules stain differentially with 
Alcian blue. Whereas reaction to Maxwell's Alcian blue tends, in most 
species, to correspond with results observed with PAS (that is, staining 
within the basal portion is less than that observed in the neck region), a 
simple exposure to aqueous Alcian blue (Steedman's procedure) revealed 
the reverse situation in Natalus, Plecotus, Pizonyx, and Molossus. In these 
genera the basal area stained more intensely than the neck. The cardiac 
glands of Sturnira yielded no positive staining with either procedure; those 



Studies of Stomachs of Selected American Bats 697 

of Artibeus stained weakly and equally throughout the tubule with both 
procedures. The aqueous reaction (Steedman's) for the remaining seven 
species was moderate to strong, and either equal in strength throughout 
the tubule, or less in intensity in the basal area (seen also in these seven 
species using Maxwell's procedure). The pattern of overall relative re- 
sponse to Alcian blue between cardiac glands of different species was roughly 
equal between the two procedures, although Steedman's was consistently 
slightly more intensive than Maxwell's. 

In summary, the results of staining cardiac glands with Steedman's 
Alcian blue revealed a pattern of distribution into one of three categories, 
depending upon the presence or absence of reversed differential staining 
within the tubular glands of the cardia. One group that had differential 
staining with the aqueous Alcian blue procedure reversed to that of PAS 
is composed of Natalus, Plecotus, Pizonyx, and Molossus. A second group, 
composed only of Sturnira and Artibeus, showed little or no reactivity 
throughout the cardiac tubules. The last group included all remaining 
species in which the cardiac glands reacted moderately to strongly, either 
in a pattern similar to that of PAS (basal staining less than upper portion) 
or with uniform staining throughout the length of the tubule. 

Results for Alcian blue staining of pyloric glands were somewhat similar 
to those obtained in the cardiac mucosa. Reversed (to that of PAS) differ- 
ential staining occurred after exposure to Steedman's Alcian blue procedure 
in three (Natalus, Plecotus, Pizonyx) of the four species in which this was 
demonstrated for the cardiac glands (and with increased intensity relative 
to Maxwell's procedure, also seen in the cardiac glands). The basal portions 
of the pyloric glands of these three species stained moderately to intensely, 
whereas the necks gave only a weak reaction (Molossus revealed only 
slight staining throughout the pyloric glands). Using Maxwell's procedure, 
basal portions of the pyloric glands of these species were consistently un- 
stained and reactions in the necks were weak to moderate. Sturnira re- 
vealed no staining of pyloric glands with either procedure; Artibeus stained 
slightly in the neck region, but only when Maxwell's procedure was used. 

None of the remaining species showed reversed differential staining 
in the pyloric glands. Except for Glossophaga and Carollia, the pyloric 
glands of which reacted uniformly and strongly only to Steedman's pro- 
cedure (and with complete lack of staining after periodic acid oxidation), 
there is a clear trend within remaining genera studied toward more 
staining of mucus using Maxwell's procedure. Rhynchonycteris, Pteroi 
Phyllostomus, and Desmodus reacted only with Maxwell's procedure, 
of the species of the Emballonuridae, Noctilionidae or 
demonstrated the reverse differential staining discussed pr 

In summary, it may be said that the cardiac and pyloric gl 



698 The University Science Bulletin 

considerably more similar histochemically among species in the Natalidae, 
Vespertilionidae and Molossidae, than among the Emballonuridae, Noc- 
tilionidae, Desmodontidae and Phyllostomatidae, with the exception of the 
uniformly poor reactivity of the phyllostomatids Sturnira and Artibeus. The 
results included here would seem to refute contentions by some workers 
that pyloric glands and glands of Brunner may be equivalent gastrointestinal 
entities. 

The mucous neck cells of the fundic mucosa of Rhynchonycteris, Noctilio, 
and species of the Phyllostomatidae (excluding Phyllostomus, Sturnira and 
Artibeus) stained weakly to intensely with Steedman's procedure, but 
stained only slightly or not at all with Maxwell's. Phyllostomus stained 
strongly with both procedures, whereas Sturnira revealed no staining in the 
mucous neck cells and Artibeus showed only a trace of staining with 
Steedman's method. The mucous neck cells of Desmodus did not react to 
Alcian blue. The remaining four species of the Natalidae, Vespertilionidae, 
and Molossidae revealed an interesting pattern of coloration. The mucous 
neck cells of Natalus reacted moderately to Maxwell's procedure, but did 
not stain with a simple aqueous method. In contrast, those of Plecotus, 
Pizonyx and Molossus all stained moderately with the aqueous procedure; 
the first and last species showed a trace of staining with Maxwell's pro- 
cedure but Pizonyx did not. 

In comparing the results obtained with PAS and Alcian blue, the most 
notable immediate contrast is that observed in the glands of Brunner. 
Although Lillie (1%5) recorded these glands as staining only red in a 
combined PAS-Alcian blue procedure, suggesting negative results in terms 
of Alcian blue, Spicer (I960) reported that although these glands in mice 
and rats react as indicated by Lillie, Brunner's glands of guinea pigs and 
rabbits yielded a moderate to strong reaction to an aqueous solution of 
Alcian blue. Carollia appears to represent an example of the latter condition 
among bats. 

It is apparent that variation exists among species of bats in the presence 
or absence and degree of reactivity to Alcian blue in various mucous cells, 
hence in relative amounts of acid mucosubstances, and is substantially 
greater than the reactivity of relatively neutral materials stained with SchifT's 
reagent. The fact Alcian blue at pH 7.1 may show greater specificity in 
reactivity with various substances than does SchifT's reagent may be a con- 
tributing factor to this variability. 

Summary. — In summary, acid and neutral mucopolysaccharides of epi- 
thelial, gastric and Brunner's gland secretions react variably to a greater 
or lesser degree to four histochemical procedures that demonstrate the 
presence of mucin materials. Variability in reactivity of gastric and duo- 
denal mucins to toluidine blue is extreme among carnivorous species; 



Studies of Stomachs of Selected American Bats 699 

herbivores and a single sanguivore are uniformly unreactive in all but 
goblet cells mucins. The results of PAS and Alcian blue procedures reveal 
interesting patterns of staining in certain portions of the gastric mucosa 
that suggest unique combinations of various mucin materials within closely 
related families or subfamilies of bats, and reveal evidence that bats with 
markedly different food preferences have different qualitative and quanti- 
tative mucopolysaccharide complements. 

Mucin Histochemistry of the Stomach in Relation to Food Habits 

The utilization of four histochemical procedures, thought to be specific 
for complex mucopolysaccharide substances, has revealed interesting trends 
and similar patterns of reactivity among certain phylogenetically and nu- 
tritionally related groups of bats. 

The most extensive gastric mucous production of any bat examined 
was that of Phyllostomus discolor. Mucins of this species reacted consistently 
and with high intensity to all three procedures employed, and these results 
reflect the presence of a particularly well-developed gastric mucosa in this 
bat. The omnivorous diet and seemingly ravenous feeding habits of this 
species evidently require abundant surface mucus to replenish a rapid loss 
of large quantities in escaping chyme. The omnivorous habits of Phyllo- 
stomus probably are related to comparatively extensive secretion of meta- 
chromatically staining gastric mucus and to equally extensive neutral epi- 
thelial mucins. In addition to the extensive mucous material in the stomach 
of Phyllostomus, as revealed by tests with PAS, the strongly positive results 
from both Alcian blue procedures indicate that chemical diversity of acid 
mucins may be more extensive in this bat than in most other species. 

Histochemical results indicate that the three frugivorous species and the 
sangivorous Desmodus are comparatively low in production of gastric 
mucins, including surface as well as deeper mucous cells. It is clear, how- 
ever, that Carollia far exceeds Sturnira and Artibeus in production of gastric 
mucus, and approaches the moderate condition of Glossophaga. Moderate 
reactions to PAS were recorded for the surface epithelium of Artibeus and 
Sturnira, although the quantity of mucus present was low for each. To- 
luidine blue and Alcian blue reacted weakly in these two species. 

In all bats but the frugivores, the surface epithelial mucin demonstrated 
by Alcian blue was found to decrease progressively in reactivity through 
the terminal portion to the pyloric sphincter; this epithelium stained con- 
sistently to the sphincter in the obligate fruit-eaters. This suggests the 
possibility of more widespread secretion of mucin in the stomachs of the 
latter group. 

In overall mucous production, the primarily nectivorous Glossophaga 
tends to resemble insectivorous and carnivorous species more than frugi- 



700 The University Science Bulletin 

vores, although it is notably similar to Carollia in several features. Mucous 
material, as revealed by PAS and Alcian blue, is more abundant throughout 
the stomach of Glossophaga than the three frugivorous kinds, although less 
abundant than in Pteronotus or Phyllo stomas. Glossophaga also reveals 
greater diversity of cell types yielding a demonstrable metachromasia than 
do any of the frugivorous species (diversity also is the tendency in insecti- 
vorous species in this regard). Glossophaga and Carollia are similar and 
unique, however, in the greater intensity of staining of the gastric foveolae 
within the fundic caecum in contrast to the remainder of the fundic mucosa 
(absent in other species, even in those with deepened gastric pits in the 
fundic caecum). Additionally, Glossophaga shows intensive staining of 
the lower mucous neck cells, otherwise restricted to insectivorous and 
carnivorous species, and this, in combination with the moderate reactivity 
of parietal cells to PAS noted above, clearly suggests the presence of an 
intermediate condition relevant in mucin production in this species. It is 
of interest to note that the parietal cells of the three frugivorous species 
(also Glossophaga) react moderately strongly with PAS, but there is no 
comparable result in any purely insectivorous or carnivorous species. The 
possibility exists, although admittedly remote, that the parietal cells of 
these frugivores may assume a mucus-producing function in the absence 
of the surface epithelial mucus found to occur in most other species of bats, 
in addition to their function as the probable source of hydrochloric acid. 

It has been fairly well established that mucous production is slight in 
the stomach of Desmodiis, in surface and deeper mucous material, as well 
as in secretions from the glands of Brunner. In the absence of numerous 
parietal and chief cells, there is little probable need for surface mucus as a 
protective device against autodigestion, and the non-abrasive nature of the * 
primary food source of this species would suggest little need for epithelial 
mucus as a lubricant against mechanical injury. 

Most insectivorous and carnivorous species of bats yielded moderate to 
strong reactions (PAS) for neutral epithelial mucins, and for acid muco- 
polysaccharides as demonstrated by Alcian blue. As regards overall mucous 
production, relatively little information emerges, using the general tech- 
niques employed in this study, to suggest differences between insectivorous 
and piscivorous species. However, the mucins of the cardiac and pyloric 
glands of Pizonyx revealed particularly strong metachromasia with toluidine 
blue, and the epithelial mucins in these two glandular areas may be gen- 
erally thought of as uncommonly abundant. In comparison to its relative, 
Plecotus, and from what can be gathered from the work of Rouk (1968) 
on two other vespertilionid bats, the cardiac and pyloric glands of Pizonyx 
are morphologically and physiologically distinctive elements in this bat, 
and for reasons that are unclear provide particularly abundant mucus at 
the two gastric orifices. 



Studies of Stomachs of Selected American Bats 701 

Some considerable variation exists in reactivity of the deeper cells of 
the cardiac and pyloric glands to particular procedures among insectivorous 
species, particularly for the two Alcian blue techniques. Because this vari- 
ation appears to have greater systematic than morphologic application, a 
discussion of it is deferred to a later section. Insectivorous, piscivorous, and 
nectivorous species reveal heavier complements of gastric mucus in the neck 
cells of the cardiac and pyloric glands than do obligate fruit-eaters and the 
sangivorous Desmodus. 

The unique presence, among the bats studied, of PAS-reactive zymogenic 
granules in the stomach of Noctilio is not readily explainable. 

Toluidine blue metachromasia is considered by most, including Spicer 
(1960, 1963), to demonstrate the presence of sulfated acid mucopolysac- 
charides. This belief is supported by autoradiographic analyses of 3-J S ab- 
sorption in the gastric mucosa of several species of rodents and one rabbit 
as presented by Jennings and Florey, 1956; these show general conformity in 
;! '"'S04 = uptake in areas correspondingly demonstrating toluidine blue meta- 
chromasia. Metachromasia may occur in some neutral mucopolysaccharides 
such as those of the glands of Brunner but Wislocki et al. (1947) suggested 
that hyaluronic acid may account for some metachromasia in tissue sec- 
tions. Tests for hyaluronic acid were negative in all species examined sug- 
gesting a lack of this substance in the gastric mucus of bats. 

Reactivity to toluidine blue O in the form of observable metachromatic 
staining reveals extensive variability between species and families, and 
among carnivorous species generally. Non-carnivorous species within the 
Phyllostomatidae tend to show less reactivity than do the carnivorous 
members examined. A principally nectivorous phyllostomatid, Glossophaga, 
which apparently often takes insects in the diet, had greater diversity of 
areas showing positive reactivity (as in the strictly carnivorous types) than 
did the three frugivorous phyllostomatids. The two species examined of 
the Vespertilionidae were notably similar, although not identical, in pattern 
of metachromatic staining. 

The glands of Brunner, although not an element of the stomach in 
most species, are morphologically and physiologically closely associated 
with the stomach, and therefore were examined histochemically. The glands 
of Brunner of all species revealed some PAS-positive material, whereas 
toluidine blue gave mixed results, and only Carollia yielded a positive 
staining with Alcian blue. It would appear that the glands of Brunner of 
Carollia, although extremely limited in distribution, are unusually active 
in the secretion of an abundance of neutral, as well as acid mucopolysac- 
charides. 

Species variability in reactivity of Brunner's glands to PAS has been 
shown to exist in rodents by Spicer (1960). He noted that Brunner's 



702 The University Science Bulletin 

glands of albino mice stained weakly, those of rats stained moderately, and 
those of guinea pigs stained more intensely than either rats or mice. The 
physiological significance of such variability is unknown. It is of interest 
to note that Lillie (1965:511) recorded the glands of Brunner as negative 
to Alcian blue, and it would appear that Carollia and perhaps Sturnira and 
Artibeus may represent genera within subfamilies of bats unusually adapted 
for mucous production by the glands of Brunner. Relatively neutral mucin 
material appears to be most consistently represented in the Brunner's glands 
of the insectivorous or partially insectivorous Phyllostomatidae, Noctilioni- 
dae, and Molossidae. The results of PAS, as noted previously, confirm the 
presence of Brunner's glands in the stomachs of Sturnira and Molossus. 

GASTRIC MORPHOLOGY IN RELATION TO FEEDING 

Food Habits of Species Examined 

Relatively little as yet is known concerning the specific nature and 
geographic variability of food habits for most species of bats. Therefore, 
we can only generalize regarding feeding preferences of certain species. 

Little or no information appears available regarding the specifics of food 
habits of Rhynchonycteris naso or, indeed, emballonurid bats in general. 
Goodwin and Greenhall (1961) noted that R. naso on Trinidad and To- 
bago, along with all other species of sac-winged bats on these islands, feed 
exclusively on insects. Perhaps mostly small, soft-bodied prey species are 
taken by this bat and are macerated to the point that identification is 
difficult. 

Noctiho labialis, unlike the fish-eating N. leporinus, appears to be an 
obligate insect-feeder as noted by Bloedel (1955) and Hooper and Brown 
(1968). According to Hooper and Brown, beetles, Lepidoptera, and bugs 
comprise a large portion of the diet of N. labialis. Numerous species of the 
Phyllostomatidae, long thought to be almost completely herbivorous or 
nectivorous, are now known to be in part insectivorous or carnivorous 
as a result of recent food habit analyses. A number of kinds including 
Pteronotus parnellii are as yet poorly studied. P. parnellii has been reported 
to be an obligate insect-feeder. Phyllostomus discolor was reported by 
Goodwin and Greenhall (1961) to eat only fruit, but Arata et al. (1967) 
observed insect remains in the stomach of P. discolor, noting that the genus 
Phyllostomus is omnivorous. The omnivorous habits of a closely related 
species, P. hastatus, are well documented (Dunn, 1933; Arata et al., 1967). 
Glossophaga soricina is a nectar-feeder, but feeds additionally on soft fruits 
(Arata et al., op. cit.) and is known also to take large insects, thus dis- 
counting the hypothesis that some insects are taken only incidently with 
ingested fruit. Although Carollia perspicillata is primarily a consumer of 
soft, pulpy fruit, Arata et al. (1967) noted that the stomach contents of 



Studies of Stomachs of Selected American Bats 703 

more than a fifth of the specimens of this species examined by them 
contained quantities of insect material. Also noted was a single instance 
of either cannibalism or predation on another bat. It would appear, on the 
basis of present information, that C. perspicillata is not truly an obligate 
fruit-eater, although it probably is less an omnivore than Glossophaga or 
Phyllostomus. Sturnira lilium is clearly an obligate fruit-eater. Artibeus 
lituratits may occasionally take some insects in the diet, although it con- 
sumes mostly soft fruits, apparently in great quantities, as food material 
passes through the digestive tract quite rapidly (in 15 to 20 minutes as 
reported by Goodwin and Greenhall, 1961). 

The three vampires of the Desmodontidae are generally thought of as 
being obligate blood-feeders, with the exception of occasional instances of 
insect-feeding by Desmodus. Considering the marked structural and be- 
havioral specialization for blood-feeding in this group, relative obligation 
to sanguivorous habits appears sufficiently verified. Wimsatt and Guerriere 
(1962) noted the rather large daily blood meals taken by Desmodus in 
captivity, averaging over 15 ml per day over a period of a year, and 
Wimsatt (1969) recently confirmed these observations in the field. 

Three of the remaining four species (Natalus, Plecotus, and Molossus) 
may be considered obligate insect-feeders. Ross (1967) noted that P. town- 
sendii is especially fond of small Lepidoptera. According to Pine (1969), 
Molossus ater in Costa Rica takes many small insects, especially ants, other 
small Hymenoptera, and a variety of beetles. Little information is available 
regarding the habits of the Natalidae; Goodwin and Greenhall (1961) 
stated only that the diet of Natalus is restricted to insects. 

Pizonyx vivesi is presently recognized as a regular fish-eater, although 
insects and small crustaceans have been observed in quantity in the stomachs 
of several specimens (Reeder and Norris, 1954). Small sardines and an- 
chovies may comprise a substantial portion of the fish diet of Pizonyx. 

It is of interest to note that many species of a number of families, al- 
though principally restricted to a non-insect diet, occasionally take insects. 

Gross Morphological Features 

Although variation in general morphology among the species of bats 
examined is extensive and each species is clearly unique in one or more 
ways, it is apparent that trends in similarity do exist among both closely 
related genera and among those with similar food habits. Several broad 
categories of general structure can be seen in the 13 species examined. 

The first of these groups, which includes Noctilio, Pteronotus, Natalus, 
Plecotus, and Pizonyx, is characterized by a stomach that is relatively short 
in relation to breadth, and that has a convex dorsal surface with no dilation 
of the dorsal surface of the fundic caecum, and is generally symmetrical, 
with the terminal portion (that region distal to the gastroesophageal junc- 



704 The University Science Bulletin 

tion) equal (or nearly so) in length to that of the fundic caecum. Also, 
the terminal stomach is relatively short and never sharply recurved craniad, 
and the esophagus enters the stomach approximately midway along the 
lesser curvature, never at a marked angle to the stomach. Two additional 
vespertilionids, Myotis velifer and Antrozous pallid us (both insect-feeders) 
examined by Rouk (1968), would seem to fit this category. This simplified, 
tubular gastric form is restricted to, although not universal among, carniv- 
orous species. Pizonyx vivesi deviates most from the basic plan discussed, 
because its esophageal entrance is slightly angled and its stomach is con- 
siderably more rounded than in the other four genera. The apparent spe- 
cializations in gross morphology observed in Piznoyx cannot be attributed 
necessarily to fish-eating, however. 

Generally speaking, the stomachs represented by this first group of bats 
closely resemble those of the Insectivora discussed by Myrcha (1967) and 
by Allison (1948). This similarity is greatest between bats (excluding 
Natalus) and hedgehogs of the family Erinaceidae, with which there is 
striking correspondence in configuration and relationship of the fundic, 
pyloric, and gastroesophageal areas, as well as similar distribution of gastric 
mucosa. Natalus, with its relatively straight pyloric tube, most closely 
resembles soricids. It is suggested that this first category of species is repre- 
sented by those with stomachs that topographically and grossly resemble 
the probable simple primitive condition in bats, species with a more spe- 
cialized fundic caecum and an elongate terminal portion being most ad- 
vanced. This statement is not intended to infer that the stomachs are 
necessarily primitive in all characteristics, but only that the basically simple 
"insectivore-type" prevails. 

Two species of obligate insectivorous bats included in this study clearly 
do not conform to the category described above. While distinctively tubu- 
lar and convex throughout on the dorsal surface, the stomachs of Rhyncho- 
nycterts naso and Molossus ater have notably elongate terminal portions in 
comparison to those of the five species previously discussed, as well as a 
sharply angled entrance of the esophagus into the stomach. Pyloric recurva- 
ture is moderate in Rhynchonycteris and pronounced in Molossus; regarding 
topography, these two species represent a second category of more special- 
ized insectivorous kinds. Possibly they are adapted for consumption of 
relatively large quantities of food during a single feeding, resulting in 
extended exposure to proteolytic digestion in the elongated pyloric portion 
of the stomach. It will also be noted that the transitional zone from fundic 
to pyloric mucosa is most extensive in these two species among all those 
examined with elongated terminal portions. This condition would tend to 
expose the food bolus to HCl-producing mucosa for a greater distance than 
in other species. A more complete analysis of quantities of food consumed 
is required to verify this suggestion. 



Studies of Stomachs of Selected American Bats 705 

A third group, which includes Curollia perspiallata, Sturnira lilium, 
and Artibeus lituratus, have stomachs with a moderately to extensively 
developed cardiac vestibule, well-developed fundic caecum that is dilated 
and hooked on the dorsal surface, and a narrow, elongate terminal portion 
that is recurved sharply craniad. Notable related specializations of the 
stomach wall are discussed in preceding sections. The conditions prevailing 
in these species would seem to relate to accommodating large quantities of 
bulky food material, consistent with the frugivorous habits of these species. 
The specializations of, and general increase in size of, the fundic caecum 
in frugivorous bats would seem to parallel a similar trend noted by Voron- 
tzov (1960) in passing from primarily albuminous (animal) to principally 
cellular (plant) feeding in muroid rodents. A chamber-like cardiac vesti- 
bule, especially well developed in Sturnira and Artibeus, increases the avail- 
able surface to which ingested material is exposed in the gastric phase of 
digestion. It is not possible at this time to conclude whether the vestibule 
is to be considered histogenetically a portion of the esophagus or a dilated 
portion of the lesser curvature, although in light of the well-developed 
gastric mucosa within the vestibule the latter explanation seems the more 
probable. As noted previously, Curollia perspicillata may be a less obligate 
fruit-eater than either Sturnira or Artibeus and some features of gross 
morphology that tend to distinguish Sturnira and Artibeus from remaining 
kinds are, indeed, partially reduced in Carolha. The cardiac vestibule is 
distinctive as a conical ampule externally in Carolha, although noticeably 
shorter and relatively narrower than in either of the other two species. The 
fundic caecum is well developed and dilated dorsally, although somewhat 
more rounded as in Glossophaga, a species to be discussed shortly. The 
pyloric sphincter of Carollia combines features observed in Artibeus (such 
as being long and symmetrical) with a structural feature of insectivorous 
or partially insectivorous kinds in which the sphincter valves increase in 
breadth toward the inner limit (narrowly tapered in Artibeus and reduced 
to a vestige in Sturnira) and terminate in a broadly rounded apex. Carolha 
clearly is a species that normally consumes mostly pulpy plant material and 
fruit juices, but perhaps has turned to such habits only relatively recently 
inasmuch as its stomach appears intermediate in many ways between 
insectivorous and frugivorous types. 

The fourth category of stomachs represented by those of Phyllostomus 
discolor and Glossophaga soricina suggests an additional type of inter- 
mediacy between strictly herbivorous and carnivorous conditions. The 
cardiac vestibule is either reduced to a vestige (Phyllostomus) or absent 
(Glossophaga), and, whereas the fundic caecum is dilated and expanded 
on the dorsal surfaces, the caecum appears relatively smaller in comparison 
to the remaining stomach than in the two obligate fruit-eaters discussed 



706 The University Science Bulletin 

previously. Only short endpieces of the terminal portions are recurved 
cranially in both Phyllostomus and Glossophaga, although the terminal 
tubular stomach is extensive in both. Also, whereas the pyloric sphincter 
retains its narrowness, as observed in other phyllostomatids, the valve in 
these two species is asymmetrical (larger on the greater than the lesser 
curvature). It is suggested that asymmetry of the pyloric sphincter is in 
some way related to carnivorous feeding in bats (additional discussion on 
this point follows in a more detailed comparison of the pyloric sphincters 
examined). In addition, the stomachs of Glossophaga and Phyllostomus 
appear to be "intermediate" in degree of symmetry, revealing more gradual 
antero-posterior curvature than observed in Carollia, Sturnira or Artibeus. 
The significance of such general arching of the stomach is obscure, although 
an hypothesis regarding the function of sharp recurvature of the pylorus is 
presented below. 

In Sturnira and Artibeus, which presumably consume rather large quan- 
tities of bulky plant material, mechanisms to retard gastric emptying, thus 
exposing food material that is difficult to digest to longer periods of gastric 
enzyme exposure, would theoretically be of selective advantage. Such 
mechanisms are well known in other kinds of herbivorous mammals (for 
example, ruminate and semi-ruminate stomachs of bovines and many herbiv- 
orous rodents). A sharply angled pyloric tube such as found in Artibeus, 
Sturnira or Carollia likely would have a distinctive crook at the point of 
recurvature when the stomach is filled to the point of distention. This 
crook in the terminal stomach might well help to retard the rapid transport 
of food through the stomach. A single insectivorous species, Molossus ater, 
was found to have a similarly sharp cranial recurvature, although the 
terminal recurved endpiece is extremely short, and the degree of constric- 
tion of the pylorus upon complete distention of the stomach with food 
is probably not great. It is apparent that the degree of recurvature of the 
pyloric portion of the stomach is to some degree a function of angulation 
of the stomach with respect to a transverse plane through the animal (that 
is, the more anteriorly displaced is the fundic caecum, generally, the greater 
the pyloric bend). In the case of the extremely large and elongate stomach 
of Molossus ater, apparent compensation for size has been made by a shift 
of the stomach up and to the left, underneath the left lobe of the liver. 
Thus, the degree of pyloric recurvature is also, most probably, a function 
of relative size of the stomach; large stomachs may be the result of in- 
creased nutrient requirements or reduced energy per unit of food or both. 
Members of the family Molossidae are strong, fast fliers, which represent 
the most advanced development of the forelimb for flight among bats. 
Superior development of flight allows at least one species, Tadarida bra- 
sihensis, to migrate, and the relatively large size of the stomach in this 



Studies of Stomachs of Selected American Bats 707 

species (Rouk, 1968), to accommodate relatively large amounts of food, 
is not surprising in this regard. 

It would appear then, that elongation of the stomach is related to in- 
creased food intake in the evolutionary history of the Chiroptera. An 
interesting parallel in this regard can be found in the Insectivora. Myrcha 
(1967) has found that the stomachs of some insectivores, notably Soricidae 
and Talpidae, reveal elongation of the terminal portion of the stomach, 
likely also in response to increased food requirements. 

The last category to be considered, containing only Desmodus, repre- 
sents the vertex of specialization in gross gastric morphology in bats thus 
far examined. It is difficult to observe any major feature of the stomach 
of Desmodus that closely resembles the condition in any other species thus 
far studied (excluding, of course, Diphylla as discussed by Grasse, 1955). 
The elongate and tubular fundus is clearly adapted to hold relatively large 
amounts of blood. Extremely elastic bands of thin, loose, bundled, muscula- 
ture and distinctive complements of loose areolar connective tissue allow 
extensive expansion of the fundic caecum. The extremely close proximity 
of the cardiac and pyloric orifices (Fig. 33) suggests, in the absence of 
apparent structural modifications of the cardia to assure passage of ingested 
blood to the left into the caecum, that the pyloric sphincter must be at least 
partially functional, or as suggested by Selkurt (1966), localized contrac- 
tions of the proximal area of the duodenum may provide the means of 
gastric closure at the lower end. The pyloric sphincter is extremely thin, 
symmetrical, and appears to be structurally most similar to that in several 
phyllostomatids, especially Carollia and Artibeus; in Desmodus it is reduced 
secondarily to a thin flap. 

The long, thin, tubular nature of the stomach of Desmodus undoubtedly 
serves to facilitate utilization of a wholly liquid meal in several ways. 
Elongation of the stomach coupled with extensive vascularization of the 
submucosa provides increased surface area for the rapid absorption and 
excretion of fluids as noted by Wimsatt and Guerriere (1962). Modifica- 
tions of the gastric mucosa, to be discussed later, may also provide for 
increased surface area. The tubular stomach, when folded on itself midway 
along its length and displaced longitudinally or transversely within the body 
cavity, would provide a mechanism for balancing the extremely large 
blood meal equally on the left and right sides of the bat, providing for 
more stable flight from the host after feeding. Vampires have been ob- 
served to engorge themselves to the point that they are unable to take 
flight after feeding (Goodwin and Greenhall, 1961); a means to provide a 
"balanced load" would seem to be advantageous to these bats. Studies 
of the intravisceral configurations of the stomach of Desmodus just after 
feeding should be carried out to test this suggestion. 



708 The University Science Bulletin- 

Table 2. Classification of pyloric sphincters of several species of bats. 

Valve flaps absent Sturnira 

Valve flaps present, broad (apex blunt) 

short, nearly symmetrical Pizonyx 

short, asymmetrical Natalus 

long, asymmetrical Noctilio 

Plecotus 

Molossus 

Rhynchonycteris 

Valve flaps present, narrow 

short, symmetrical Desmodits 

long, symmetrical Carollia 

extremely long, tapering, symmetrical Irtibctis 

long, asymmetrical Pteronotus 

Phyllostomus 
Glossophaga 

Extensive variation in morphology of the pyloric sphincter in bats (Table 
2) presents an interesting physio-morphological problem. Without excep- 
tion, every species of carnivorous or partially carnivorous bat examined in 
this study revealed some degree of asymmetry of the pyloric sphincter 
(that is, the valve was not uniform in length or breadth, or both, through- 
out the circumference of the gastroduodenal junction, the larger portion of 
the valve usually being on the greater curvature). This asymmetrical con- 
dition is in contrast to the pyloric sphincters of herbivorous and sanguivo- 
rous species, which, although highly variable in length among species, are 
equal throughout their circumference within a given bat. Dearden (1969) 
examined the pyloric sphincters of four genera of microtine rodents, noting 
two types of asymmetry in the sphincter — that in which the larger sphinc- 
teric mass was on the greater curvature, and conversely, that in which the 
valve was larger on the lesser arc. Dearden suggested that in those species 
of voles with the greater mass of muscle on the greater curvature, sphincteric 
action would seem to be of a milking nature, whereas symmetrically circu- 
lar closure would likely result from the other condition. Although a well- 
developed membrane of gastric mucosa comprises an important part of the 
pyloric sphincter in bats, which lack the pediculated squamous flaps ap- 
parently important to the closure mechanism in some species of voles, the 
anatomical variability observed in pyloric sphincters in bats suggests that 
these mammals also have functional differences in gastroduodenal closure 
mechanisms. A relatively effective gastroduodenal closure mechanism, such 
as would theoretically be produced in those species examined with sym- 
metrical sphincters (Sturnira, Artibeus, Carollia, Desmodits), would pre- 
sumably be of greatest advantage in bats feeding on materials that are 
difficult to assimilate, such as pulpy plants and blood. Nothing is known 



Studies of Stomachs of Selected American Bats 709 

regarding the physiology of the pyloric sphincter in bats, although condi- 
tions in primarily non-carnivorous bats suggest that additional studies on 
pyloric sphincter morphology and physiology in bats might prove most 
interesting in relation to feeding ecology. 

Fundic Caecum 

Comparative examination of the fundic caeca of bats reveals little spe- 
cialization that can be directly related to feeding habits. An exception is 
Desmodus, in which the entire stomach is essentially a thin-walled caecum 
adapted to accept and retain a meal of blood that requires no physical 
breakdown; the complex and normally thick muscular wall observed in 
fundic caeca of other bats is lacking. No species examined lacked a fundic 
caecum, and all species (excluding Desmodus) revealed fundic caeca that 
were but slight alterations of a basic plan. All species of the Phyllostomati- 
dae, with the possible exception of Pteronotus, possess a fundic caecum 
dilated at the apex and hooked dorsally to a greater or lesser degree. This 
feature is restricted to phyllostomatids; it is best developed in Artibeus, 
Sturnira and Curollia, to a lesser degree in Glossophaga. It is reasonable 
to suggest that this condition is also related to increasing effective gastric 
volume in those species that consume large quantities of plant bulk and 
liquid material, being absent (or reduced) in those kinds that have retained 
(or returned to) insectivorous and carnivorous habits. The fundic caecum 
of all carnivorous species consists of a simple, usually short, blind, lateral 
extension of the mid-stomach, excepting the slight terminally hooked apex 
observed in that of Phyllostomits (an omnivore), and functions to hold a 
portion of the food material during the period of gastric churning. 

Cardiac Vestibule 

The cardiac vestibule and its probable functional significance in three 
fruit-feeding or at least predominantly fruit-feeding kinds (Sturnira, Arti- 
beus, and Carolha) has previously been discussed. The vestibule in these 
three species is unique among those examined in that it represents a nearly 
distinct gastric entity. It is of interest that fruit-eating megachiropteran 
bats may have an equally well-developed cardiac vestibule, implying further 
a relationship between consumption of large complements of fruity material 
and invasion of gastric mucosa craniad into the lower portion of the 
abdominal esophagus. All insectivorous and partially carnivorous species, 
as well as the one sanguivorous species studied, have either a relatively 
short vestibule or lack one entirely, with the possible exception of Natalus 
stramineus. The vestibule of Natalus is rather extensive in breadth, although 
not as distinctively delimited from the lesser curvature as the vestibule of 
the three frugivorous species. Although nutritional requirements of in- 
sectivorous and carnivorous bats likely are equivalent to, or greater than, 



710 The University Science Bulletin 

those of fruit-eaters, usable caloric substrate is likely greater per weight 
of food material ingested. Insect-eating bats macerate their food into fine 
fragments, and the few carnivorous kinds whose digestive efficiency has 
been examined have been noted to leave little recognizable animal material 
other than matted hair and hard fragments in the fecal material, even 
when whole mice are consumed. This efficiency in the digestive process 
suggests relatively lower requirements of food mass than in fruit- and blood- 
feeders; hence, comparatively less modification of gastric proportions are 
required to increase internal volume and surface area. It is of interest to 
note that Natahts, the single insect-feeder possessing a large cardiac vesti- 
bule, also has a relatively small fundic caecum and a terminal stomach 
that is moderate in length, but extremely narrow. It is probable that the 
enlarged cardiac portion in this species has developed in response to an 
otherwise rather limited gastric volume. 

Relatively little is known concerning the mechanism of gastroesophageal 
closure in bats, which is surprising in light of their interesting and almost 
unique habit among mammals of hanging upside down while roosting 
(after and sometimes even during the period of feeding). An anatomically 
distinctive cardiac valve is unknown in bats with the exception of that 
reported herein for the genus Noctilio. This structure evidently has been 
derived independently in this genus, perhaps in response to the acquisition 
of fish-eating habits later abandoned by N. /tibialis. The stomachs of both 
species of noctilionids appear comparatively small in relation to body 
size, and the unique cardiac valve, assumed functional to some degree as 
specialization of the mucularis mucosae is apparent, may prevent back-up 
of large quantities of ingested fish from a relatively small stomach. Should 
a cardiac closure mechanism exist in other bats, as seems likely, it must 
be represented by a physiologically specialized muscular component of 
circular fibers and muscularis mucosae at or near the gastroesophageal 
junction, perhaps consisting partly of locally thickened musculature as in 
those species with a prominent incisura cardiaca. Anatomical specialization 
of muscle layers at the gastroesophageal junction in bats other than Noctilio 
is restricted to a slight breakup of fibers of the muscularis mucosae, the 
significance of which is unclear, and some notable local thickening of the 
circular muscle layer just below the gastroesophageal junction in Rhyncho- 
nycteris, Glossop/nigu, and Molossus. Botha (1958a) examined the mucosal 
folds in the cardia of an unspecified species of bat, probably an insectivorous 
vespertilionid, and concluded that the mucosa-lined ridges at the orifice 
formed a watertight papilla, effectively closing the gastroesophageal junc- 
tion. It is possible that those species with local thickening of the circular 
muscle layer in this area {Rhynchonycteris , Glossophaga, and Molossus) have 
efficient closure of the stomach, although no physiological evidence is avail- 



Studies of Stomachs of Selected American Bats 711 

able to support this point. It will be noted that the three species of frugi- 
vores all have extremely thin musculature at the gastroesophageal junction 
and throughout the cardiac vestibule. This condition does not, of course, 
dispel the possibility of the existence of an active closure mechanism, 
although the strength and efficiency of a sphincter may be more limited 
in these species. 

Musculature 

The walls of all stomachs examined were composed of an outer longi- 
tudinal and inner circular layer, and two extremely thin intra-submucosal 
sheets of smooth muscle, a pattern which is common to other kinds of 
mammalian stomachs. Variability between species in corresponding depths 
of various portions of the stomach frequently was found to be less than 
variation within a single stomach, especially between tubular and saccular 
portions of several species. 

Most of the species examined revealed some degree of increased thick- 
ening of the greater curvature wall, in contrast to that of the lesser curva- 
ture. The relative amount of, or specific region of, the greater curvature that 
shows this condition is, again, quite variable and rarely expressed through- 
out all of the greater arc, or within both outer layers. In Phyllostomus , 
Natalus, and Molossus, the muscular wall is consistently thicker through 
the greater arc, by virtue of the circular layer, than most of the correspond- 
ing wall of the lesser curvature. This condition generally holds for 
Pteronotus, although the terminal portion is only slightly thicker in the 
greater curvature than in the lesser (hence uniformly shallower than in 
the lesser arc in the three species mentioned above). The circular layer 
of the greater curvature wall in Noctilio is distinctive in being only 
slightly thicker than the lesser wall, whereas the longitudinal layer is 
markedly increased in depth in the medial and terminal portions of the 
greater curvature. 

The stomachs of Glossophaga, Carollia, Sturnira, and Artibeus all pos- 
sess some degree of sacculation of the fundic caecum, with a greater or 
lesser degree of reduction of the muscular wall within this portion of the 
stomach. The musculature of the greater curvature of the fundic caecum 
is thicker than the facing wall on the lesser curvature in Glossophaga and 
Carollia, whereas these two surfaces are essentially equal in depth and thin 
throughout in Sturnira and Artibeus. The terminal pyloric portions in these 
four bats are always thicker than the proximal wall and generally thickest 
on the greater curvature, although the difference is not noticeable in 
Carollia. 

The functional significance of much of this irregularity in the depth 
of muscle layers is unclear, although conjectures regarding answers to such 
questions follow where appropriate. In the absence of any pertinent physio- 



712 The University Science Bulletin 

logical information concerning gastrointestinal tracts of bats, it is difficult 
to explain such extensive variability. 

It would seem likely that differential muscle development that results 
in greater masses of circular fibers on the greater curvature would likely be 
a response to increased food intake, which would demand additional basal 
support in the gastric sac to accommodate the increase in food mass. Al- 
though marked modifications in this condition of differential musculature 
are found to exist in localized portions of the stomach in many species, 
most still express the condition, at least in the terminal portion of the 
stomach. Exceptions to this trend, although rare, do exist and much addi- 
tional work is needed to clarify the role of these conditions. 

Especially in Plecotus and less so in Pizonyx, some marked uniformity 
in depths of corresponding portions of the lesser and greater curvatures is 
apparent. Musculature of the fundic caecum in each of these two species 
is moderate and uniform in depth, except for reduction at the extreme 
apex; thickening in the pylorus is pronounced in Piznoyx and slight in 
Plecotus. Plecotus and Rhynchonycteris most closely approach the condition 
of equal thickness of the muscle wall throughout the stomach among 
the carnivorous and herbivorous species examined. The muscular wall in 
the stomach of Rhynchonycteris is remarkably uniform, Plecotus somewhat 
less so; all remaining species show some local thickening, usually in the 
pyloric endpiece. It is suggested that the condition of nearly uniform depth 
of the stomach musculature in these two species is a relatively generalized 
condition. 

Decrease in depth of the caecal wall is apparent among herbivorous kinds, 
whereas this region remains deep or even increases in depth over other 
areas in obligate insectivores and carnivores. It would seem likely that 
gastric volume is of greater importance in herbivorous kinds, hence loss 
of musculature in the caecum permits an increase in elasticity. On the 
other hand, prominent musculature in the proximal portion is of greater 
advantage to insectivorous kinds in increasing the area over which large, 
rythmical contractions take place, hence the physical mechanism of gastric 
mixing. It is interesting that Ph\llostomus discolor, a probably omnivore, 
approaches the condition of equal muscular depth throughout the stomach 
to a greater degree than any other phyllostomatid examined, with only 
slight and nearly symmetrical thickening in the terminal portion of the 
stomach: this suggests a retention of generalization in this feature in 
Phyllostomus and perhaps in other phyllostomatines. 

Quite localized decrease in caecal musculature at the extreme apex 
(owing to decrease in depth of the circular layer), as occurs in Rhynchonyc- 
teris, Noctilio, Phyllostomus, Plecotus, and Piznoyx, probably reflects a 
compromise to obtain increasing flexibility and yet remain the substantial 
musculature required for gastric mixing in carnivorous bats. 



Studies of Stomachs of Selected American Bats 713 

Asymmetrical pyloric thickening, discussed previously, is noted to be 
widespread among bats examined. Where present this does not necessarily 
correspond with, nor is it restricted to, those kinds with similar asymmetry 
of the nearby pyloric sphincter, nor is it correlated with relative size of the 
sphincter. The significance of this feature remains obscure, and its wide- 
spread occurrence suggests that studies of gastric motility and the physio- 
logical role of the pyloric sphincter in bats need be undertaken to help 
resolve these questions. 

As regards the outer musculature in Desmodus, the extensively bundled 
fibers within both outer layers provide a suitable mechanism to explain 
prominent expansion of the stomach to hold the large quantities of blood 
taken in individual feedings. Some suggestion of increased depth of the 
greater curvature over that of the lesser curvature is to be found near the 
gastroesophageal junction in Desmodus, the significance of which, as in 
other bats revealing this feature, remains obscure. Although the bundles 
of circular fibers are variable in size, the total depth of the outer wall is 
fairly uniform throughout the fundic caecum in Desmodus, suggesting 
symmetrical expansion of the caecum upon filling. The mechanism to 
allow extensive expansion is not, apparently, reduction of muscular tissue 
in depth, but an increase in the elastic portion of the wall by way of in- 
creased complements of areolar connective tissue. 

An index for determining differences and degree of variability in ar- 
rangement of circular muscle fibers was taken to be the extent of bundling 
of this layer as viewed in a mid-longitudinal section of the stomach. This 
condition, as observed in mid-longitudinal section, is restricted for the 
most part to the lateral and lower wall of the fundic caecum, but in some 
is present up to the medial area of the greater curvature. There are, 
however, several notable exceptions to this trend. The stomach of Desmodus, 
as discussed previously, illustrates extensive bundling throughout all of the 
fundic caecum, unique in this regard among bats examined. Most species 
examined had some degree of bundling of the circular layer in the fundic 
caecum, usually in the lower portion and within the greater curvature. 
The only exception to this trend was found in Natalus, in which the 
bundling present in a mid-longitudinal section was a small area just 
proximal to the pyloric bend. The function of gathering of circular fibers into 
bundles appears to be a gain in degree of elasticity of the stomach, expressed 
to some degree in the fundic caecum or greater curvature (or both) of most 
species, especially the phyllostomatids and Molossus. The apparent lack of 
this feature in the fundic caecum of Natalus is particularly interesting in 
light of the unique feature in this bat of increased muscular depth within 
the fundic caecum in comparison to the remainder of the stomach. This 
suggests that the stomach of Natalus is relatively non-plastic, with little 



714 The University Science Bulletin 

capacity for expandability of the fundic portion. Whether or not this condi- 
tion restricts the relative amount of food consumed in comparison to other 
species cannot at this time be concluded, but seems unlikely to be the case 
in light of the rather extensive cardiac vestibule near the esophagus in 
Natalus, providing some increase in gastric volume. 

Relatively little reference has thus far been made to the longitudinal 
layer of the stomach, because it consists of only a thin sheet (in comparison 
to the overlying circular complement) throughout most portions of stomachs 
of nearly all of the 13 species studied. However, in at least three species of 
insectivorous bats {Pteronotus, Noctilio, Natalus), representing three dis- 
tinctive families, the longitudinal layer is prominently developed locally 
throughout part or all of the fundic caecum. In the case of Natalus it is 
as deep as, or deeper than, the thick circular layer throughout most of the 
caecum. Where such hyper-development occurs, the longitudinal layer of 
the caecum becomes several times the depth of the remaining portion of 
this layer, which as noted is usually extremely thin. Except in Pteronotus, 
the omnivorous Phyllostornus discolor reveals the best development of the 
longitudinal layer among the primarily frugivorous phyllostomatid group, 
and some correlation between this condition and carnivorous habits in bats 
seems apparent. An increase in longitudinal muscle fibers would increase 
the strength of the shortening component for a rhythmical-segmentation 
contraction, which would presumably increase the effectiveness of mixing 
caecal contents with that of the midportion of the stomach. This suggestion 
must, of course, be tested by examining additional species with similar 
feeding habits, and examining the mode of gastric motility in these mam- 
mals. 

Gastric Mucosa and the Glands of Brunner 

orientation of folds 

In general, only two patterns of directional orientation of mucosal folds 
were observed in the 13 species examined. The first condition, that of 
longitudinal orientation of parallel folds, was found in the tubular portions 
of all stomachs irrespective of the feeding type represented. Longitudinal 
orientation to a greater or lesser degree was also observed in the fundic 
caeca and cardiac vestibules of all insectivorous species, the omnivorous 
Phyllostornus, the piscivorous Pizonyx, and sanguivorous Desmodus. Within 
the non-tubular portions of stomachs of several species (including Phyllo- 
stornus, Glossophaga, Plecotus, Pizonyx, Molossus, and Desmodus), which 
have mostly longitudinal orientation of folds, there occurs slight to moderate 
wrinkling (or zig-zag arrangement) of the rugae, although no true inter- 
digitation of the folds is to be found. This condition is lacking in several 
other carnivorous species, most of which (Rhynchonycteris excepted) also 
reveal relatively thick musculature in the fundic caecum. The condition 



Studies of Stomachs of Selected American Bats 715 

of wrinkling of the rugal folds appears related to increased ability to 
stretch the caecal portion of the stomach. Wrinkling of the mucosal folds 
provides for additional muscular support in the form of folded sheets of 
muscularis mucosae to compensate for the extremely dilated outer muscu- 
lature that would result upon marked distention of the stomach. It is not 
surprising to find that this feature is lacking in those species with deep 
caecal musculature, which probably restricts the degree of expansion of 
the caecum. 

The second condition of orientation of mucosal folds is that of trans- 
verse anterior-posterior orientation. This situation occurs 1) infrequently 
among a few species of insectivorous kinds (here found restricted to 
infrequent, short transverse folds in brief areas of the fundic caecum, 
usually the upper wall), 2) in the fundic caecum of Glossophaga and 
Carollia as narrow transverse ridges along the sides, and 3) in an extreme 
condition of highly complex, heavily interdigitating ridges in the non- 
tubular fundus of Artibeus and to a lesser degree that of Sturnira. A possible 
function of these interdigitating folds, that of helping to secure food ma- 
terials within the fundic caecum for extended periods of time, has been 
previously discussed for Artibeus. An alternative, or at least supplementary, 
hypothesis would be to suggest that these folds provide for increased elas- 
ticity in addition to the already thin outer wall of the fundic caecum to 
assist in accommodation of large meals of fruit such as are known to be 
taken by Artibeus and Sturnira. 

LOCAL REDUCTIONS IN MUCOSAL DEPTH 

Two conditions of local modification in mucosal depth in comparison 
to the remainder of the stomach seem worthy of mention, as they are 
widespread among the species examined. In the stomachs of all species 
studied, except for Carollia and Sturnira, the gastric mucosa is clearly 
deeper in the lesser than in the greater curvature. This condition does not 
appear correlated with muscular depth as it is found to exist in species 
with an hypertrophied wall along the lesser arc and along the greater arc, 
as well as in species with uniform or nearly uniform muscular depth, and 
is usually restricted to within the zone of fundic mucosa. Perhaps no rela- 
tionship exists between mucosal depth and digestive efficiency or strength 
of the gastric wall, although this seems unlikely. Evidence that the condi- 
tion is best developed and, therefore, most important in carnivorous kinds 
is hardly conclusive, and thorough study of these patterns should be under- 
taken in other kinds of mammals with diverse food habits. 

A second condition is that of reduced depth of the mucosa within the 
fundic caecum in comparison to the remainder of the stomach wall. Gen- 
eralized reduction throughout the fundic caecum occurs in only one species, 
Glossophaga soricina, whereas localized reduction only at the lateral apex 



716 The University Science Bulletin 

of the caecum (usually slight) occurs in the stomachs of Rhynchonycteris, 
Phyllo stomas, Pkcotas, Pizonyx, and Molossus. Reduced glandular depth 
in the caecum is generally the result of reduction in number of all types of 
cells present, but particularly the zymogenic cells. It would appear that 
this condition is again part of the general hypertrophy observed in other 
layers at the apex of the caecum in these species, related to increased local 
elasticity. The herbivorous species and the insectivorous and carnivorous 
species with thick musculature within the caecum do not demonstrate this 
condition of local reduction of mucosal depth. They have other mechanisms 
for increasing the volume of the caecum or have caeca that are relatively 
non-elastic. 

distributions of types of gastric mucosa 

Extensive variation between species and surprisingly little variation 
within species was found in distribution of most of the types of gastric 
mucosa. The zone of cardiac glands at the gastroesophageal junction is a 
relatively narrow band in all species examined, but approaches a condition 
of being a distinctive element of the mucosa in Pizonyx vivesi, unequaled 
in this feature by any other bat examined. The cardiac glands of Pizonyx 
are extremely large and occupy the entirety of the short cardiac vestibule. 
These glands are composed entirely of mucus-producing cells, and are pre- 
sumed to be a source of lubricating substance applied to the food material 
as it enters the stomach (Bensley, 1903). All species examined possessed at 
least a few of these glands, but their extensive development in Pizonyx 
suggests that the food consumed by this species requires an especially 
abundant coat of mucus. Whether the fish consumed by Pizonyx present 
in especially abrasive mass for the stomach is unknown. It is of some in- 
terest to reflect on the fact that the greatest degree of specialization at the 
gastroesophageal junction among bats is to be found in two fish-eating 
kinds studied, the other being Noctilio leporinus. It is indeed apparent, 
however, that these specializations are of a wholly different order — hyper- 
development of cardiac glands in Pizonyx and an anatomical cardiac 
sphincter lined with squamous epithelium in Noctilio. Although seemingly 
unrelated anatomically, it can be argued that these specializations produce 
similar conditions at the entrance of the esophagus into the stomach, that 
of a smooth or heavily mucus-coated surface. The import of this feature 
as it relates to fish-eating habits should be investigated further. 

Although not unique in spatial distribution within the stomach, the 
cardiac glands of Natalus are large in size and are unusual in that not many 
occur beneath the stratified squamous epithelium at the extreme upper 
limit of the cardiac vestibule. These glands are nearly identical morpho- 
logically to (and about the same size as) the glands of Brunner in the 
duodenum of this bat. These two types of glands are also identical in 



Studies of Stomachs of Selected American Bats 717 

reactivity to two histochemical procedures for mucins (unique among bats 
examined). The unique similarity in the two glandular types in Natalus 
should be investigated further. Study of the ultrastructire of these cells, 
along with those of the glands of Brunner, might provide additional evi- 
dence regarding the apparent parallelism in structure, as well as shed light 
on their poorly understood physiological roles in the mammalian stomach. 
The pyloric glands are easily distinguished from the cardiac glands in 
Natalus and the marked similarity of the latter to Brunner's glands is in- 
deed unusual. 

Myrcha (1967) examined the distribution of zones of gastric mucosa 
in 27 species of insectivores representing three families. He concluded that 
the often-observed elongation of the terminal portion in many species was 
a direct result of increased food consumption, and that an extension of the 
transitional zone between fundic and pyloric mucosa was a consequence 
of this elongation. Among the bats studied, general correspondence was 
found between elongation of the terminal portion and elongation of the 
zone of transitional mucosa except in the Noctilionidae and Phyllostomati- 
dae. In the latter groups the pattern not only breaks down but shows some 
reversals of the trend among the included species. 

An example of this breakdown may be seen in two obligate insectivorous 
bats, Pteronotus parnellii and Noctilio labialis. Both have relatively short 
terminal portions of the stomach, yet have extremely long transitional 
zones, apparently developed at the expense of the pyloric mucosa. This 
pattern becomes further complicated upon examination of the transitional 
areas of Phyllostomus and Artibeus, two species with somewhat elongate 
terminal portions, yet extremely narrow transition zones. The distribution 
of pyloric mucosa may be an important controlling influence in determining 
the amount of pylo-fundic transitional mucosa present in bats, because 
those species with extremely broad transitional zones (for example, Rhyn- 
chonycteris, Noctilio, and Pteronotus) have small complements of pyloric 
mucosa, and, conversely, those species with extremely narrow transitional 
zones (for example, Phyllostomus, Artibeus, and Pizonyx) have rather 
extensive regions of pyloric mucosa, as compared with their closest rela- 
tives. It is thus apparent that the phenomenon of coelongation of the 
terminal stomach and zone of transitional mucosa, as outlined by Myrcha 
(1967) for the Insectivora, is not found among the frugivorous Micro- 
chiroptera. Specific information concerning the relationship of feeding habits 
to the physiological significance of the pyloric mucosa is needed to help 
resolve this question. 

The gastric glands of bats are, for the most part, similar to those of 
other kinds of mammals in types of cells present and in the distributional 
order of cells within the tubules from one end to the other. However, the 



718 The University Science Bulletin 

proportion of individual tubules occupied by various types of cells, and 
the depth of various mucosal and submucosal layers differ from the arrange- 
ment in many other mammals. The gastric glands generally may be re- 
garded as shallow in comparison to those of rodents and insectivores, 
animals of similar size. The cause or effect of the shallow nature of these 
glands appears primarily to be a loss of parietal and mucous neck cells in 
medial areas of the tubules. The result of this loss is greater equality in 
length among regions occupied by zymogenic, parietal, and mucous neck 
cells in bats, because the zones of parietal and mucous neck cells tend to 
be much longer in rodents, insectivores, and many other mammals with a 
mucosal membrane. In addition, there is a corresponding reduction in the 
amount of submucosal connective tissue throughout much of the gastro- 
intestinal tract of bats, resulting in a presumably highly elastic intestine 
and equally elastic local areas of the stomach. 

As regards the distribution of pyloric mucosa, no clear-cut patterns 
emerge with respect to similarities among species with similar food prefer- 
ences. It is evident however, that frugivorous leaf-nosed bats, as well as 
the omnivorous Phyllostomus, tend to have extensive complements of these 
glands, with rather narrow transitional zones. This condition, although 
much less apparent, is found also in Pizonyx. Only Natalus among the 
insectivorous bats examined has pyloric glands in an unusual length of the 
stomach. Because pyloric glands histogenetically represent a type of gastric 
mucosa in which the cells responsible for secretion of materials engaged in, 
or complementary to, proteolytic breakdown are replaced by mucous pro- 
ducing types, it is understandable that these highly active digestive cells 
are less widespread in distribution in frugivorous bats. Several species of 
carnivorous bats (Rhynchonycteris, Pteronotus, and Nocttlio) reveal a near 
loss of true pyloric glands with marked proximal expansion of parietal cells 
toward the pyloric sphincter. A consequence of the elongation of the 
stomach in frugivorous species has been an extension of the pyloric zone 
only, rather than corresponding expansion of the transitional zone as seen 
in most insectivorous kinds. Phyllostomus has acquired a stomach con- 
figuration and muscular wall more like those of carnivores, and yet, has 
retained the zonation of gastric mucosa most often observed in frugivorous 
species. The rather extensive distribution of pyloric mucosa in Natalus, 
the stomach of which is unusual in many features, is unique among in- 
sectivorous species examined, a probable consequence of notable elongation 
of the terminal portion. Glossophaga, an apparent "intermediate" in the 
true sense of the term, more closely approaches the "insectivorous" condi- 
tion in distribution of pyloric and fundic mucosa than does Phyllostomus. 

These conclusions do not refute the possibility that relatively numerous 
pyloric glands at the proximal end of the stomach may be of some con- 



Studies of Stomachs of Selected American Bats 719 

siderable physiological advantage in one or more of the bats studied, par- 
ticularly the frugivorous species in which they are so abundant. Numerous 
physiological roles have been suggested for the mucous producing cells of 
these glands, including their being sites of anti-autodigestion enzymes and 
neutralizes for pepsin among other things. However, until the exact physio- 
logical role of these glands is understood, it is difficult to draw meaningful 
conclusions regarding their relative abundance in the stomachs of bats. 

Fundic Mucosa 

The fundic mucosa comprises the most morphologically specialized and 
presumably the most physiologically specialized portion of the gastric 
mucosa in bats. Fundic glands always occupy the majority of the internal 
surface area of the stomach and are characterized by the presence of both 
parietal and zymogenic cells within the tubules. That portion of the gastric 
mucosa that contains numerous parietal cells but no zymogenic cells is here 
not considered fundic, but rather transitional mucosa, mostly for the sake 
of convenience. It is therefore to be assumed that when distributions of 
parietal cells in the stomach as a whole are considered, the transitional 
mucosa is included as a part of these distributions. 

The zymogenic cells are considered to be the site of secretion of the 
enzyme pepsin, activated to proteolytic action in the presence of HC1. 
Although the greatest numbers of chief cells were found in several insec- 
tivorous species, other insectivorous and carnivorous bats revealed smaller 
complements of chief cells than did frugivorous species. There is no clear 
relationship, therefore, between food habits and relative numbers of zymo- 
genic cells. The substantial reduction in the number of these cells in 
Desmodus is taken to be a consequence of general hypertrophy of fundic 
glands in this bat. The striking depletion of zymogenic cells and especially 
parietal cells in Desmodus elicits interesting questions concerning the ability 
of Desmodus to assimilate food material, particularly in light of the high 
percentage of solids found to be utilized metabolically by Desmodus (Wim- 
satt and Guerriere, 1962). 

It is clear that wherever zymogenic cells are notably abundant, the dis- 
tribution of parietal cells is correspondingly limited to a narrow portion 
of the tubule. The relative proportion of the gastric tubule occupied by 
chief cells seems to be a stronger controlling force afTecting the numbers 
of parietal cells present than the varieties of foods taken in the diet. Frugiv- 
orous, insectivorous, and piscivorous species were observed with both re- 
stricted and extensive complements of parietal cells in the fundic and transi- 
tional areas. Such extensive variation remains unexplained. Allison (1948) 
has suggested that distribution and numbers of parietal cells may remain 
static in the gastric mucosa of insectivores, and Rouk (1968) observed no dif- 



720 The University Science Bulletin 

ferences in the distributions of parietal cells in active and hibernating Myotis 
velifer. These facts suggest that differences between species in complements 
of at least some cell types in the gastric mucosa is, in all likelihood, real and 
definable interspecific variation, the significance of which requires addi- 
tional investigation for adequate explanation. 

It will be noted that in three species examined (Rhynchonycteris, Noctilio, 
and Pteronotits), nearly all of the gastric mucosa contains some parietal 
cells and these cells are, for the most part, extremely small and numerous. 
Perhaps large numbers and extensive distribution of parietal cells in these 
species compensates for small cell size to provide the needed concentration 
of HC1. 

As in the case of zymogenic cells in bats, no clear-cut patterns emerge 
with regard to relative abundances of parietal cells in comparison to types 
of foods taken. The especially numerous parietal cells found to occur in 
Artibeus may be related in some way to the rapid food transport in this 
species, with a substantial flow of HCl being required to counterbalance 
the presumed short exposure to digestive fluids. This conclusion is only a 
supposition and requires further analysis. It may be that parietal cell 
abundance is related to rate of food passage in other species, or that the 
abundance of other cell types in the gastric mucosa is also related to time 
of exposure of the food to gastric digestion, although little information is 
available on this subject. 

With the exception of Artibeus, in which the parietal cells are both 
abundant and of large size, there is a trend among the bats examined to 
suggest that these cells are relatively large in species in which they are few 
in number; conversely, the cells are smaller in those bats in which they are 
more numerous. To my knowledge, no such suggestion of an inverse 
correlation between size and abundance of parietal cells is to be found 
for any other kind of mammal. It is of interest to note that Allison (1948) 
observed the parietal cells of species of the Insectivora to be progressively 
smaller from the lower to the upper boundary of their distribution within 
an individual fundic tubule, a condition found to occur only in Rhynchonyc- 
teris among the 13 species of bats examined herein. At least 10 of the 
remaining 12 species (the condition is unclear in Carollia and Desmodus) 
revealed the reverse situation — the largest cells were toward the surface 
of the mucosa. The parietal cells closest to the surface of the gastric lining 
are presumed to be the "oldest." Differences in growth capacity and physio- 
logical activity in parietal cells in various species of bats, as well as between 
major subgroups of mammals, seems probable. 

It is apparent that greater local variation within the stomach in distri- 
bution of parietal cells and zymogenic cells, as well as in depth of gastric 
pits, is found among insectivorous or partially insectivorous bats as a group, 



Studies of Stomachs of Selected American Bats 721 

but not among frugivorous, piscivorous, or sanguivorous kinds. The fundic 
mucosa of insectivorous kinds is consistently better developed in the lesser 
curvature than in the greater curvature; in terms of development longitu- 
dinally it usually reaches its maximum within the middle portion, decreas- 
ing in richness of cell types in either direction laterally. The reduction 
within or at the apex of the fundic caecum in many species appears also 
to be related to increasing the elasticity of the stomach wall. 

Few comments have thus far been made specifically regarding relative 
amounts of gastric mucus produced via the combined secretions of pyloric 
and cardiac glands, surface epithelium, cells of the gastric pits, and the 
deeper mucous neck cells. Considerable variation exists among species in 
depth of the surface epithelial layer, as well as depths of the gastric pits 
within the fundic mucosa. Those species with relatively deep gastric pits 
or particularly thick surface epithelium might theoretically produce the 
most epithelial mucus. The three primarily frugivorous species of the 
Phyllostomatidae along with Phyllostomus have both extensive pyloric 
mucosa and relatively deep surface epithelium, suggesting some probable 
advantage to abundant mucous secretion in relation to frugivorous habits. 
The surface epithelium of Glossophaga is comparable in depth to most 
insectivorous species. The extremely well-developed gastric epithelium of 
Desmodus may be traceable to its close phylogenetic relationship with the 
Phyllostomatidae; however, it is most likely related in some way to the 
extensive capacity for absorption of water from the stomach into the 
vascular network, as comparatively little surface mucus was observed in 
Desmodus in histochemical tests for mucopolysaccharides. 

The gastric pits of the fundic tubules, formed by an apparent involution 
of the surface epithelium, are unusually well developed in Phyllostomus. 
This feature, in combination with the extensive complement of deep, mas- 
sive pyloric glands and notably thick surface epithelium, presents Phyllo- 
stomus as an unusually heavy producer of gastric mucus. The extensive 
quantities of food known to be consumed in rather short periods of time by 
species of Phyllostomus would demand high levels of enzyme and acid. 
Rather complete decomposition of materials such as bone (Dunn, 1933) 
must produce gastric environments that threaten the mucosal lining unless 
a deep mucous coat is present. Finally, as regards gastric mucous produc- 
tion, several species had differential increase in the depth of the gastric 
pits within the fundic caecum in comparison to the remainder of the 
stomach. This increase is most often accompanied by corresponding re- 
duction in numbers of other cell types in these glands; the advantage of 
having relatively more surface mucus in this area than in the remainder 
of the fundic portion is unclear. 

There exists a trend among certain groups of mammals for a relatively 



722 The University Science Bulletin 

greater distribution of the glands of Brunner in the upper duodenum of 
herbivorous species as opposed to carnivorous species. However, an im- 
pressive group of exceptions to the trend may be found without exhaustive 
search, as in bats. One general statement regarding the distribution of 
these submucosal glands in bats can be made: these highly convoluted 
elements are always restricted to the immediate vicinity of the pyloric 
sphincter, usually in a small mass, with little or no representation in the 
form of smaller, isolated clumps inferior to the main mass, as is common 
in other kinds of mammals. Some variation exists, however, in relative 
distribution of these glands among bats, in the morphology of glands, and 
in at least two species in degree of abundance in the terminal portion of 
the stomach. Two of the three frugivorous species, Artibeits and Carollia, 
have Brunner's glands so reduced in distribution as to be nearly lost; those 
of Sturnira are only moderate in abundance in the duodenum, although 
they are extensive in the pyloric stomach proximal to the pyloric sphincter 
(shared apparently only with Molossus). Among the phyllostomatids ex- 
amined, the greatest masses of Brunner's glands were observed in Pteronotus 
and Phyllostowus, two species that likely consume insects in quantity. 
As the functional role of secretions of Brunner's glands remains obscure, 
it is difficult to comment on the meaning of variation in their distribution. 
It would appear that certain frugivorous kinds have secondarily lost most 
of these glands. Carleton (1935) noted that the location of the common 
bile or pancreatic duct in some mammals may determine the limit of dis- 
tribution of Brunner's glands into the duodenum, although there are excep- 
tions. The common bile duct enters the duodenum only a few millimeters 
distal to the pyloric sphincter in bats, and may be related in fact somehow 
to distribution of the glands of Brunner. However, it must be noted that 
in Molossus the glands occur slightly beyond the entrance of the common 
bile duct into the duodenum, and, in fact, an ampulla is formed at the 
junction as a result of invasion of Brunner's glands into the duct. The 
general consensus now is that the glands of Brunner function to protect 
the lining of the initial portion of the duodenum from the damaging 
action of acid material escaping from the stomach (Grossman, 1958). In 
this respect, the near loss of Brunner's glands in a species such as Artibeus 
with its rather extensive number of large parietal cells is indeed perplexing. 
The invasion of the glands into the pyloric portion of at least two distantly 
related species, the food habits of which are entirely different, suggests 
that this condition might be widespread among certain families. Probably 
it confers additional protection against damage to the pyloric mucosa of 
Molossus and also especially to Sturnira, although the particular need for 
such a mechanism in these species is not readily apparent. 

In concluding the discussion of gastric structure in relation to feeding, 



Studies of Stomachs of Selected American Bats 723 

it seems appropriate to suggest that control over morphology may be two- 
fold, one factor affecting the general shape and relative proportions of 
various regions of the gastric sac and the other controlling the distribution 
and abundance of elements of the gastric mucosa. Each factor acts sepa- 
rately in gastrointestinal evolution to accommodate particular types or 
quantities of food. In this regard, it appears that the quantity of food 
consumed may be an especially important criterion in the evolution of 
gastric structure. Elongation of the stomach has been a particularly im- 
portant event in the acquisition of frugivorous habits in bats, whereas most 
insectivorous and all piscivorous species examined have retained a relatively 
simplified stomach. Those species thought to be non-obligate feeders and 
that take substantial quantities of both animal and plant matter, were 
found to be somewhat intermediate in the structure of the stomach. Gastric 
structure as observed in bats does not usually correspond with the conditions 
found to exist in the Insectivora, although Natalus and Rhynchonycteris 
closely parallel the Insectivora in several features. In general, evolution 
away from insectivorous feeding in bats has resulted in noteworthy spe- 
cialization in one or more structural features of the stomach. 

SYSTEMATIC RELATIONSHIPS AS REVEALED BY 
GASTROINTESTINAL STUDIES 

Although the stomach of each of the 13 species examined was clearly 
distinctive in one or more major features, clear-cut relationships were 
found within families as well as among closely related families. Modifica- 
tion of gastric structure to accommodate and assimilate foods other than 
insects has led to marked changes in external configuration of the stomach. 
Upon further investigation, however, various features of internal gastric 
structure emerge to reveal more precisely, relationships among various 
groups of bats. Because the Phyllostomatidae represents the family with 
the most genera examined, and also has the greatest intrafamilial variability 
in food habits of included species, it presents the most interesting body 
of information. 

Relationships of bats with the Insectivora, both phylogenetic and as 
regards food habits, bring to mind consideration of which species of bats, 
if any, seem to conform most closely to gastric structure as presently known 
for the Insectivora. The stomach of Natalus stramineus, as noted previously, 
closely resembles those of soricid insectivores in stomach configuration, dis- 
tribution of mucosal elements, and particularly in external and internal 
morphology of the terminal or tubular portion. In addition, the stomachs 
of Rhynchonycteris and Plecotus also are generalized externally and internally 
and are strikingly like those of several talpid and erinaceid insectivores, 
respectively. These relationships suggest a generalized condition in early 



724 The University Science Bulletin 

chiropterans, most, if not all, of which were probably insect-feeders. The 
specializations observed in other insectivorous bats such as Molossus and 
Noctilio are modifications of the relatively primitive plan seen in the Ves- 
pertilionidae, Natalidae, and Emballonuridae. 

Within the Vespertilionidae, all species thus far examined reveal marked 
symmetry of the stomach (esophageal junction located nearly midway along 
the lesser curvature) and a short, slightly recurved pyloric tube. The list 
of examined species showing these features includes Myotis vehjer, Plecotus 
townsendii (plus the unnamed species of Plecotus discussed by Fischer, 
1909), Antrozous pallidas, and Pizonyx vivesi. Eisentraut (1950) included 
a drawing of the stomach of Eptesicus serotinus that, although appearing 
somewhat distended as do most of the stomachs illustrated by that author, 
is in general correspondence with those of other vespertilionids. Pizonyx 
and Antrozous represent the greatest deviants from the basic plan of overall 
stomach configuration. The stomachs of both are more rounded than those 
of other vespertilionids, and Pizonyx differs further in its atypical distribu- 
tion of mucosal zones. Hence, Pizonyx is even less like Myotis, with which 
it recently has been merged on the basis of karotypic evidence (Baker and 
Patton, 1967) than the other vespertilionid genera that have been studied. 

The stomach of Natalus stramineus is much like that of Plecotus town- 
sendii in general topography and in distributions of zones of gastric mucosa, 
but differs in degree of elongation of the pyloric tube and cardiac vestibule. 
It is difficult to determine whether similarity in structure is due to parallel 
adaptation to similar feeding characteristics or close phylogenetic association, 
or both. It has been suggested that the Natalidae may be systematically 
closely associated with the vespertilionids through the genus Kerwoula 
(see Dobson, 1878, for a discussion of relationships), and the results of 
gastrointestinal studies support, to some degree, the contention of close 
relationship between the two families. 

Although highly variable among themselves, the genera of the Phyllo- 
stomatidae examined have several features in common that tend to dis- 
tinguish them as a group from the other families considered in this study. 
In all species examined, the muscular portion of the pyloric sphincter tends 
to be narrow in relation to size of the stomach, or may be partially lacking 
(Pteronotus) or nearly completely lost (Sturnira). An extremely narrow 
sphincter was also found to occur in Desmodus. Also among phyllostom- 
atids, with the possible exception of Pteronotus, there occurs some degree 
of dilation of the dorsolateral terminus of the fundic caecum, although this 
expansion is not restricted to the dorsal surface. 

Recent investigations of the systematic relationships of the nominal 
subfamilies Stenoderminae and Sturnirinae have led most investigators to 
the conclusion that Sturnira should be considered a stenodermine. The 



Studies of Stomachs of Selected American Bats 



725 





B 




F 

Fig. 53. A diagram illustrating the gradual increase in specialization in stomachs of some 
phyllostomatid bats. Pteronotus parnellii (A); Phyllostomus discolor (B); Glossophaga soricina 
(C); Carollia perspicillata (D); Stiirnira lilium (E); Artibeus litwatus (F). Stomachs art- 
drawn to scale. 



726 The University Science Bulletin 

results of my gastrointestinal studies strongly support the proposal that 
Sturnira should be placed in the Stenoderminae. 

In addition to the specific features that tend to group the phyllostomatids 
(including the Desmodontinae), it is also of interest that certain genera 
in the family resemble one another or a representative of another family 
in unique ways. As noted, the stomachs of Artibeus and Sturnira (Fig 53) 
are similar, resembling each other in gastric morphology to a much greater 
degree than either resembles representatives of the other subfamilies studied. 
Additionally, Phyllostomus and Glossophaga are notably similar, whereas 
Carollia is more or less intermediate between this pair and the previously 
discussed pair, possibly resembling Glossophaga more than the other genera. 
The stomach (described by Rouk, 1968) of another glossophagine bat, 
Leptonycteris sanborni, closely resembles that of Glossophaga in the structure 
of the terminal portion and differs generally only in the size of the cardiac 
vestibule (larger in Leptonycteris). Although each is specialized in some 
ways, Pteronotits and Noctilio are strikingly similar in several features, 
particularly the distributions of parietal and zymogenic cells, morphology 
of the gastric glands, and general external topography, suggesting the possi- 
bility of closer systematic relationship than has been previously supposed. 

A systematic revision of the subfamily Chilonycterinae recently was 
completed by James D. Smith. The results of my study indicate that this 
group likely deserves familial recognition, complementing his findings based 
on other characters. 

It is of interest to note that if one arranges six species of phyllostomatids 
examined (excluding Desmodus), each of which represents a presently 
recognized subfamily according to phylogenetic arrangements such as those 
of Hall and Kelson (1959), Koopman (1967), or Simpson (1945), there is 
a gradual transition from insectivorous feeding (Pteronotus) to frugivorous 
habits (Artibeus) with variously omnivorous kinds interposed in sequence 
(see Fig. 53). Corresponding closely with this change in food habits is a 
progressive alteration of gastric structure from the simplified condition of 
Pteronotus to greater elongation and compartmentalization of the stomach, 
culminating in the condition found in Sturnira and Artibeus. Trends in 
other morphological features complement the trends in specialization ob- 
served in stomachs of phyllostomatid bats. The sequence presented in 
Figure 53 is not intended to imply that one species is necessarily older than 
another in an evolutionary sense, but that modifications in food habits 
from the presumed early insectivorous condition have resulted in progressive 
modifications in the morphology of the stomach as well. The desmodontines 
are clearly so highly specialized in gastric morphology that it is difficult to 
relate them to other North American groups, although close relationship 
with the Phyllostomatidae has been shown by Forman et at. (1968) using 
other characters. 



Studies of Stomachs of Selected American Bats 727 

Stomachs of molossids thus far examined reveal that this family consists 
of relatively specialized insectivorous bats with stomachs that are clearly 
different from those of other insectivorous species. The stomach of Tadarida 
brasiliensis (examined by Rouk, 1968) differs from that of Molossus ater 
only slightly, containing the same degree of elongation of the terminal 
stomach and a prominent mid-longitudinal notch in the greater curvature. 

The histochemistry of deeper cells of the cardiac and pyloric glands of 
bats reveals trends difficult to interpret in light of morphological patterns 
observed; yet the results are of interest with respect to interfamilial rela- 
tions. It will be recalled that the results of the PAS procedure showed less 
reactivity within the basal arc or bulbular portion than in the upper, neck 
region of the cardiac and pyloric glands of all species, except for an equiva- 
lent reaction from both glandular areas in the case of Molossus. When Max- 
well's Alcian blue technique was employed for acid mucopolysaccharides in 
these glands, the same pattern of differential staining as occurred with PAS 
was obtained. However, employment of the aqueous Alcian blue (Steed- 
man's method) yielded a similar pattern of results to those of PAS and 
Maxwell's in all species but Sturnira and Artibeus, which showed little 
reactivity at all, and a reversed condition (in which the basal cells in the 
cardiac and pyloric glands stained considerably darker than the upper 
cells) in a group including Natalus, Plecotus, Pizonyx, and Molossus. Al- 
though the physiological import of such results is unclear, the consistency 
of this differential staining in both cardiac and pyloric glands among repre- 
sentatives of the Natalidae, Vespertilionidae, and Molossidae supports the 
present placement of these families in the superfamily Vespertilionoidea. 
Winge (1941) suggested that the Molossidae may have arisen from primitive 
vespertilionids (based on similar structure of the femur and ulna), and 
the close relationship of the latter group with the Natalidae has been 
previously discussed. My results suggest that a broader application of more 
specific histochemical procedures upon the gastric mucosa might prove useful 
in resolving questions regarding relationships among higher taxonomic 
categories of bats. 



LITERATURE CITED 

Allison, A. C. 1948. The stomach in South African Insectivora, with notes on the organiza- 
tion of mammalian gastric glands. J. Anat. 82:249-261. 

Arata, A. A., J. B. Vaughn and M. E. Thomas. 1967. Food habits of certain Colombian bats. 
J. Mammal. 48:653-655. 

Baker, R. }., and J. L. Patton. 1967. Karyotypes and karyotypic variation of North American 
vespertilionid bats. J. Mammal. 48:270-286. 

Bensley, R. R. 1903. The structure of the glands of Brunner. The Decennial Publ., Univ. 
Chicago, 10:279. 

. 1905. The cardiac glands of mammals. Amer. J. Anat. 2:105-156. 

Bloedel, P. 1955. Hunting methods of fish-eating bats, particularly Noctilio leporinus. J. 
Mammal. 36:390-399. 



728 The University Science Bulletin 

Botha, G. S. M. 1958a. Mucosal folds at the cardia as a component of the gastrooesophageal 

closing mechanism. British J. Surg. 45:569-580. 
. 1958/'. Histological observations of the gastrooesophageal junction in the rabbit. J. 

Anat. 92:441-446. 

1958<~. A note on the comparative anatomy of the cardio-oesophageal junction. Acta 



Anat. 34:52-84. 
Carleton, A. 1935. Distribution of Brunner's glands in the duodenum of mammals. Proc. 

Zool. Soc. London, pp. 385-390. 
Clara, M. 1940. Untersuchungen iiber den farberischen Nachweis des Schleimes in den 

Driisenzellen beim menschen. Z. mikr. anat. Forsch. 47:183-246. 
Cuvier, G. 1805. Organes de la digestion. Lecons d'anatomie comparee. Paris, Volume 4 

(see pp. 31-32). 
Davenport, H. A. 1960. Histological and Histochemical Techniques. W. B. Saunders, Phila- 
delphia. xxii+401 pp. 
Dearden, L. C. 1966. Histology of the gastro-esophageal junction in certain microtine rodents. 

J. Mammal. 47:223-229. 
. 1969. Stomach and pyloric sphincter histology in certain microtine rodents. J. 

Mammal. 50:60-68. 
Dobson, G. E. 1875. Conspectus of the suborders, families, and genera of Chiroptera arranged 

according to their natural affinities. Ann. Mag. Nat. Hist. ser. 4, 16:345-357. 
. 1878. Catalogue of the Chiroptera in the . . . British Museum. Taylor and Francis, 

London, xlii+567 pp. 
Dunn, L. 1933. Observations on the carnivorous habits of the spear-nosed bat, Phyllostomns 

discolor panamensis Allen, in Panama. J. Mammal. 14:188-199. 
Eisi ntraut, M. 1950. Die Ernahrung der Fledermausen (Microchiroptera). Zool. Jahrb. Jena. 

79:114-177. 
Fischer, H. 1909. liber funktionelle Anspassungen am Fledermausmagen. Arch. ges. Physiol. 

129:113-137. 
Forman, G. L., R. J. Baker and J. D. Gerber. 1968. Comments on the systematic status of 

vampire bats (family Desmodontidae). Syst. Zool. 17:417-425. 
Glass, G. B. J. 1968. Introduction to Gastrointestinal Physiology. Prentice-Hall, Englewood 

Cliffs, N.J. 207 pp. 
Gomori, G. 1952. Histochemical studies of the gastric mucosa. J. Nat. Cancer Inst. 13: 

1039-40. 
Goodwin, G. G., and A. M. Greenhall. 1961. A review of the bats of Trinidad and Tobago. 

Bull. Amer. Mus. Nat. Hist. 112:189-301. 
Grasse, P. P. 1955. Traite de Zoologie. Mammiferes. Masson et Cie, Paris, 17:1758-1804. 
Grossman, M. I. 1958. The glands of Brunner. Physiol. Rev. 38:675-690. 
Hale, C. W. 1946. Histochemical demonstration of acid polysaccharides in animal tissues. 

Nature 157:802. 
Hall, E. R., and K. R. Kelson. 1959. The Mammals of North America. Ronald Press, New 

York. l:xxx+l-546+79. 
Hooper, E. T., and J. Brown. 1968. Foraging and breeding in two sympatric species of neo- 
tropical bats, genus Noctilio. J. Mammal. 49:310-312. 
Huxley, T. H. 1865. On the structure of the stomach in Desmodus rufus. Proc. Zool. Soc. 

London, pp. 386-390. 
Jarvi, O., and P. Lauren. 1951. On the role of heterotopias of the intestinal epithelium in 

the pathogenesis of gastric cancer. Acta Path. Microbiol. Scand. 29:26-44. 
Jennings, M. A., and H. W. Florey. 1956. Autoradiographic observations on the mucous 

cells of the stomach and intestine. Quart. J. Exp. Physiol. 41:131-152. 
Kolb, A. 1954. Biologishe Beobachtungen an Fledermausen. Saugetierkundl. Mitt. 2:15-26. 
Koopman, K. F. 1967. Bats. In Recent Mammals of the World . . . (S. Anderson and J. K. 

Jones, Jr., eds.), pp. 109-150. Ronald Press, New York. 
Langley, J. N. 1880-82. On the histology of the mammalian gastric glands, and the relation 

of pepsin to the granules of the chief cells. J. Physiol. 3:269-291. 
Leblond, C. P. 1950. Distribution of periodic acid-reactive carbohydrates in the adult rat. 

Anat. Rec. 86:1-49. 
Lillie, R. D. 1965. Histopathologic Technic and Practical Histochemistry. McGraw-Hill, 

New York, xii+775 pp. 
Mathis, J. 1928tf. Beitrag zur Kenntnis des Fledermausdarmes. Z. mik. anat. Forsch. 12: 

594-647. 
. 1928Z>. Uber die Brunnerschen Drusen and iiber Kornchenzellen einiger Fledermause. 

Anat. Anz. 65:1-7. 



Studies of Stomachs of Selected American Bats 729 

Miller, G. S. 1907. The families and genera of bats. Bull. U.S. Nat. Mus. 57:1-282. 
Mowry, R. W. 1963. The special value of methods that color both acidic and vicinal hydroxyl 

groups in the histochemical study of mucins . . . Ann. New York Acad. Sci. 106:402-423. 
Myrcha, A. 1967. Comparative studies on the morphology of the stomach in the Insectivora. 

Acta Theriol. 12:223-244. 
Owen, R. 1868. On the Anatomy of Vertebrates. Mammals. Longman, Green and Co., 

London, vol. 3:915 pp. 
Pantin, C. F. A. 1948. Notes on Microscopical Technique for Zoologists. Cambridge Press, 

New York. 79 pp. 
Park, H., and E. R. Hall. 1951. The gross anatomy of the tongues and stomachs of eight 

new world bats. Trans. Kansas Acad. Sci. 54:64-72. 
Pine, R. H. 1969. Stomach contents of a free-tailed bat, Molossus ater. J. Mammal. 50:162. 
Reeder, W. G., and K. S. Norris. 1954. Distribution, type locality, and habits of the fish- 
eating bat, Pizonyx vivesi. J. Mammal. 35:81-87. 
Robin, H. A. 1881. Recherches anatomique sur les Mammiferes de l'ordre des Chiropteres. 

Ann. Sci. Nat. Zool. 12:1-180. 
Rollett, A. 1871. Bemerkungen zur Kenntiss der Labdrusen und der Magenschlcimhaut. 

Unters Inst. Physiol. Hist., Graz. 
Ross, A. 1967. Ecological aspects of the food habits of insectivorous bats. Proc. West. Founda- 
tion Vert. Zool. 1:205-264. 
Rouk, C. S. 1968. Comparative gastric histology of selected American bats. Master's thesis, 

Oklahoma State Univ., Stillwater. 51 pp. 
Rouk, C. S., and B. P. Glass. 1970. Comparative gastric histology of American bats. J. 

Mammal. 51:455-472. 
Schofield, G. 1953. The argentaffin and mucous cells of the human intestine. Acta Anat. 

18:256-272. 
Schultz, W. 1965. Studien iiber den Magen-Darm-Kanal der Chiropteren. Ein Beitrag zum 

Problem der Homoltfisierung von abschnitten des Saugetierdams. Zeit. Wissenschaft., Zool. 

171:241-391. 
. 1970. Einige Bemerkungen zum Bau des Verdauungstraktes und der systematischen 

Stellung des Spitzzahnflughundes, Hurpionycteris ivlnteheadi Thomas, 1896 (Megachirop- 

tera). Sonderdruck auf Z. f. Siiugetierkunde. 35:81-89. 
Selkurt, E. E. (ed.) 1966. Physiology. Little, Brown and Co., Boston. 2nd ed., xii-f-796 pp. 
Simpson, G. G. 1945. The principles of classification and a classification of mammals. Bull. 

Amer. Mus. Nat. Hist. 85 :xvl+ 1-350. 
Spicer, S. S. 1960. A correlative study of the histochemical properties of rodent acid muco- 
polysaccharides. J. Histochem. Cytochem. 18:18-33. 
. 1963. Histochemical differentiation of mammalian mucopolysaccharides. Ann. New 

York Acad. Sci. 106:379-388. 
Vorontzov, N. N. 1960. The ways of food specialization and evolution of the alimentary 

system in Muroidea. Symp. Theriol. Brno. pp. 360-377. 
Walker, E. P. 1968. Mammals of the World. Johns Hopkins Press, Baltimore. 2nd ed., 

vol. 1 :xlviii+644 pp. 
Wimsatt, W. A. 1969. Transient behavior, nocturnal activity patterns, and feeding efficiency 

of vampire bats (Desmodtts rotundas) under natural conditions. J. Mammal. 50:233-244. 
Wimsatt, W. A., and A. Guerriere. 1962. Observations on the feeding capacities and ex- 
cretory functions of captive vampire bats. J. Mammal. 43:17-27. 
Winge, H. 1941. The interrelationships of the mammalian genera. C. A. Reitzels Forlag, 

Copenhagen, 1:418 pp. (translated from Danish by E. Deichmann and G. M. Allen). 
Wislocki, G. B., H. Bunting and E. W. Dempsey. 1947. Metachromasia in mammalian 

tissues and its relationship to mucopolysaccharide. Amer. J. Anat. 81:1-37. 



THE UNIVERSITY OF KANSAS 

SCIENCE BULLETIN 

Vol. XLIX Pages 731-874 April 7, 1972 No. 11 



TABLE OF CONTENTS 

ABSTRACT 733 

I INTRODUCTION 734 

MATERIALS AND METHODS 734 

ACKNOWLEDGMENTS 737 

II SKELETAL MORPHOLOGY OF PHANAEUS VINDEX 

MACLEAY 737 

The Head 

Head Capsule 738 

Appendages of the Head 746 

The Cervix 758 

The Thorax 

The Prothorax 758 

The Pterothorax '. 762 

The Pterothoracic Endoskeleton 769 

The Legs 770 

The Wings and their Articulations 772 

The Abdomen 

The Abdominal Terga 774 

The Abdominal Sterna 776 

External Genitalia 776 

III COMPARATIVE SKELETAL MORPHOLOGY 

OF PHANAEINES 776 

The Head Capsule 779 

Appendages of the Head 787 

Thoracic Segments 793 

Legs 806 

Wings - 811 

Abdomen 812 

IV ADAPTIVE SIGNIFICANCE OF ASPECTS OF 

PHANAEINE SKELETAL MORPHOLOGY 813 

V CLASSIFICATION OF THE PHANAEINES 815 

Artificial Key to the Genera of Phanaeine Scarabaeinae 819 

Diabroctis Gistel 820 

Sulcophanaeus Olsoufief? 821 

Phanaeus MacLeay 826 

Oxystemon Castlenau 835 

Coprophanaeus Olsoufieff 839 

Dendropuemon Perty - 843 

Tetramereia Klages 851 

Homalotarsus Janssens 852 

Megatharsis Waterhouse 854 

VI EVOLUTION AND PHYLOGENY OF PHANAEINES 855 

Some Preliminary Considerations 855 

Phylogenetic Speculation 857 

Generic Affinities 861 

Parallelisms 864 

Evolutionary Implications of Phanaeine Ecology and Distribution .. 867 

LITERATURE CITED 871 

APPENDIX: Material Studied 874 






FRONTISPIECE 

Phanaetis iindex MacLeay. Upper left, dorsal view of female; upper right, dorsal view of male. 
Lower left, lateral view of female; lower right, lateral view of male. (About twice natural size.) 



Comparative Skeletal Morphology, Systematics and Evolution 
of the Phanaeine Dung Beetles (Coleoptera: Scarabaeidae) 1 



w. d. Edmonds- 



abstract 



The principal purpose of this paper is a re-examination of the supraspecific 
classification of phanaeine Scarabaeinae based largely upon a comparative study 
of the skeletal morphology of adults. Part II is a detailed morphological study of 
Phcinaeus vindex MacLeay, which was chosen as a "model" species. In this part, 
which is essentially an independent study of scarabaeine morphology, a termi- 
nology is devised which is used in subsequent parts. Part III is an account of the 
observed variation in skeletal morphology among phanaeines. It is largely upon 
this variation that the revised classification is based. Part IV is a short resume 
of the presumed adaptive significance of skeletal features of phanaeine morphol- 
ogy concerned with fossorial habits. The revised supraspecific classification of 
phanaeines presented in part V incorporates several significant changes from 
older systems: 1) the phanaeine genera (not collectively referred to as the sub- 
tribe Phanaeina) include Diabroctis (-Taurocopris), Sulcophanaeus, Phanaeus, 
Oxysternon, Coprophanaeus, Dendropaemon, Tetramereia, Homalotarsus and 
Megatharsis; Oruscatus, Gromphas and Rolbites are excluded from "Phanaeina"; 

2) two names, Sulcophanaeus and Coprophanaeus, proposed by G. d'Olsoufieff 
for subgenera of Phanaeus are now used as names of genera quite different in 
composition from the subgeneric taxa to which the names originally referred; 

3) three new subgenera are proposed, Mioxysternon for some species of Oxy- 
sternon, and Paradendropaemon and Coprophanaeoides for species of Dendro- 
paemon; 4) extensive use is made of the informal categories of species-group and 
species-complex where appropriate in all larger genera. Part VI is a discussion 
of phanaeine evolution and phylogeny, including generic affinities arrived at by 
analysis of the taxonomic distribution of derived character states as well as dis- 
cussions of parallelisms and the evolutionary implications of phanaeine ecology 
and distribution. Phyletically, the phanaeine Scarabaeinae, which are presumed 
to comprise a monophyletic group, appear to consist of two distinct phyletic 
lineages (the Oxysternon-Phanaeus and Coprophanaeus-Dendropaemon genus- 
groups), together comprising the bulk of the known species, in addition to a 
number of small phyletically isolated groups representing relicts of several older 
phyletic elements. 



1 Contribution No. 1478 from the Department of Entomology, The University of Kansas, 
Lawrence, Kansas. 

2 Present address: Department (if Biological Sciences, California State Polytechnic College, 

Pomona, California 91768. 



734 The University Science Bulletin 

PART I: INTRODUCTION 

Phanaeines are perhaps one of the best generally known groups of 
Scarabaeidae. Their striking metallic colors and the often bizarre appearance 
of "horned" males have made them showy collectors' items for some 200 years. 
So well noted were they by early collectors that over 80% of all known species 
(about 90) had been described by 1875. The only serious attempt to date 
(and a generally very good one) to examine the group in toto was made by 
Gregor d'Olsoufieff (1924), who, in an extensive monograph, provided the 
first comprehensive keys to species and proposed many new supraspecific 
groupings. 

The primary intent of this paper is a re-examination of Olsoufieff's 
supraspecific classification of phanaeines based primarily upon the compara- 
tive skeletal morphology of adults. Before this could be done critically, 
however, it was necessary to examine in detail the skeletal morphology of a 
"moder species to provide a base for the comparative study. Phanaeus vtndex 
was arbitrarily selected as this model. Part II of this paper is a treatment of 
the skeletal morphology of P. vindex as well as those aspects of musculature 
necessary for interpretation of skeletal features. Here a terminology is de- 
vised which is used in other sections. Since an effort was made to compare 
P. vindex with other, non-phanaeine dung beetles, this is in effect an inde- 
pendent study of scarabaeine morphology; as such, it is hoped that it will be 
useful to other students of Scarabaeidae. 

Part III is an account of some of the observed variation of skeletal mor- 
phology; the conclusions in subsequent parts are based largely upon the data 
assembled here. Part IV is an examination of the adaptive significance of 
certain morphological features; although this deals specifically with pha- 
naeines, many of the statements there are applicable to other dung beetles as 
well. Part V is the systematic section. The concluding section (VI) is a 
discussion of the evolutionary aspects of phanaeine skeletal morphology, 
biology and distribution and the bearing of these aspects on their presumed 
phylogeny. 

MATERIALS AND METHODS 

The morphological study of Phanaeus vindex is based upon examination of 
both dried museum specimens and specimens preserved in liquid. Like material 
of Coprophanaeus lancifer, a much larger species, was studied simultaneously as 
a check on several interpretations of the morphology of P. vindex, particularly 
those dealing with muscles. All other material used in this study was dried. For 
examination of skeletal features, all dried material was first soaked overnight in 
cold 10° Cl KOH solution. In addition, it was found that the normally much 
folded hind wings could be extended for easy examination by detaching and soak- 
ing them briefly in the same solution. 

The majority of the illustrations are line drawings made mostly with the aid 



Evolution of the Phanaeine Dung Beetles 



735 



of an ocular grid. Because of the great variation in size of structures, they are not 
all drawn to the same scale. Photographs are used to illustrate those structures 
not easily depicted in ink drawings. Efficient illustration of many features requires 
several views; for orientation in these instances a large, solid dot connects guide- 
lines indicating the same point in drawings of the two or more aspects. 

Following is a list of abbreviations used in the figures and not defined in 
captions: 

dorsal portion of eye 
dorsal process of tentorium 
dorsal surface of labrum 
elytron 
. mesepimeron (anterior, lateral 
and posterior surfaces) 
metepimeron 

posterior margin of mesepimeron 
elytral epipleuron 
epipleural shelf 
mesepisternum 
metepisternum (anterior and 
lateral surfaces) 
fronto-clypeal ridge 
fronto-clypeal sulcus 
femur 

profemoral tuft 
labio-gular fimbria 
flagellum of antenna 
flexor sclerite of elytron (and 
hind wing) 
metafurcasternum 
galea 
glossa 
glossal bar 
glossal flap 
gula 

gular ridge 
gular suture 

horizontal lamella of mesonotum 
head wall 
hypostoma 

inflexed margin of basisternum 
(of presternum) 
invaginated anterior margin of 
mesepisternum 
inflexed margin of gula 
inflexed margin of sternellum 
(of presternum) 
incisor lobe of mandible 
interspiracular pad 
lateral arm of hypopharyngeal 
suspensorium 
lacinia 

labral muscles 
labial palp 
lateral clypeal carina 



abas2 


mesobasalar apophysis 


deye 


acx 


articulation of coxa 


dprt 


afr 


apical fringe of labrum 


dsula 


afur2, 3 


meso-(meta-)furcal arm 


ely 


afur2 P 


lateral process of mesofurcal arm 


epim2a, 


amcly 


anterior margin of clypeus 




ampr 


antcro-median process of labral 


epim3 




suspensorium 


epimpm 


amt 


anterior margin of tentorium 


epip 


anf 


antennifer 


epips 


ans 


antennal socket 


epis2 


ap 


aedeagal apodeme 


epis3 a ,i 


apely 


articular process of elytron 




apglb 


apodeme of glossal bar 


fcr 


apmcly 


apical margin of clypeus 


fes 


apmmi.2 


apodemes of mandibular muscles 


fe 


apst 


prosternal apophysis 


fet 


arl 


articular line (between clypeus 


fim 




and labrum) 


fla 


armx 


articulation of maxilla 


fs 


at 


acrotergite of abdominal tergum 




ata 


anterior tentorial arm 


fst 


atibc 


anterior protibial carina 


gal 


awp 


anterior metanotal wing process 


gl 


ax 


axillary sclerite 


gib 


ba 


basalis of mandible 


glf 


bas2, 3 


meso(meta-)basalare 


gu 


bfur2, 3 


base of meso(meta-)fiirca 


gur 


br 


median brace of mesofurca 


gus 


bst 


basisternum (of presternum) 


hi 


bstc 


basisternal crest 


hw 


ca 


canthal area 


hyp 


car 


cardo 


imbst 


cbr 


cephalic brachium 




cly 


clypeus 


imepis 


clypr 


clypeal process 




cmb 


comb of incisor lobe of mandible 


imgu 


conj 


cnniunctivus of mandible 


imstn 


cr 


transverse tibial carina 




cpr 


cephalic process 


inlb 


ex 


coxa 


ip 


cxd 


coxal depression of sternum 


la 


da 


dorsal arm of hypopharyngeal 






suspensorium 


lac 


dacx.3 


dorsal articulation of metacoxa 


Iami.2 


darm 


dorsal articulation of mandible 


lbpl 


darsgal 


dorsal articular sclerite of galea 


lcrcly 



736 



The University Science Bulletin 



If 
lfos 

libs 
lobe 

Ipc 

Ishpx 

mac 

mbfcs 

mbr 

mcon 

md 

mem 

mgpmt 

mlhpx 

molb 

morg 

mrhpx 

mshpx 

mt 

mth 

mxpl 

nt,,i 

oa 

ocar 

ocfor 

oclb 

ocr 

oes 

oford 

oforv 

ogs 

paar 

par 

pbs 

ped 

pfos 

phb 

plrg2, 3 

pls2, 3 

pm 

pmpr 

pmts 

pntm 

pocr 

pocs 

pol 

polb 



lateral file of labrum 

lateral pronotal fossa 

lateral labial sclerite 

postero-ventral lobe of 

metapleuron 

lateral pronotal carina 

lateral hypopharyngeal sclerite 

mandibular acetabulum 

mesal branch of fronto-clypeal 

sulcus 

median brush of labrum 

mandibular condyle 

muscle disc of metepisternum 

membrane 

median groove of prementum 

median lobe of hypopharynx 

molar lobe of mandible 

molar ridges of mandible 

median ridge of hypopharynx 

median hypopharyngeal sclerite 

mentum 

mouth 

maxillary palpus 

posterior inflection of mesonotum 



ora 



arm 



occipital area of parietals 

occipital foramen 

occipital lobe 

occipital ridge 

oesophagus 

optic foramen of dorsal portion 

of eye 

optic foramen of ventral portion 

of eye 

oculo-gular space 

paraocular area of parietals 

paraocular ridge 

pleurobasisternal suture 

pedicel 

posterior pronotal fossa 

phallobase 

meso(meta-)pleural ridge 

meso(meta-) pleural suture 

paramere 

postero-median process of labral 

suspensorium 

premental sclerite 

circumnotal ridge 

postoccipital ridge 

postoccipital suture 

postocular line 

postocular lobe 



pors paraoral sclerite 

ppha paraphragmal area of prescutum 

(of metanotum) 

ppl propleuron 

pr process 

prcar articular process of cardo 

prmx articular process of maxilla 

proc preocular carina 

pscu prescutum (of metanotum) 

pstn metapostnotum 

ptbr posterior tentorial bridge 

ptd posterior tentorial depression 

ptibc posterior carina of proti'oia 

pwp posterior metanotal wing process 

phi. 2. j phragma (first, etc.) 

rg ridge 

rmgl receptacle of mandibular glands 

scu2,3 meso(meta-)scutum 

sct2,3 meso(meta-)scutellum 

sen scnsilla 

smt submentum 

sp scape 

spfr scapal fringe 
ssi, ii, in, iv stipital sclerite I, etc. 

st stria 

stn sternellum (of prosternum) 

sts mesometasternal suture 

sub3 m <i metasubalar muscle disc 

su hpx hypopharyngeal suspensorium 

Si-8 abdominal sternum 1, etc. 

T].-s abdominal tergum 1, etc. 

tab metepisternal tab 

tar tarsus 

tcrcly transverse clypeal carina 

tib tibia 

tibs tibial spur 

tibt tibial teeth 

topr tormal process of labral 

suspensorium 

tr trochanter 

trn trochantin 

vacx3 ventral (sternal; articulation of 

metacoxa 

varm ventral articulation of mandible 

varsgal ventral articular sclerite of galea 

veye ventral portion of eye 

vl vertical lamella of mesonotum 

w w-suture of metanotum 

wp2, 3 meso(meta-)pleural wing 

process 

y articular surface of mesepimeron 

z articular surface of mesepimeron 



Evolution of the Phanaeine Dung Beetles 737 

ACKNOWLEDGMENTS 

I would like to express my thanks to Dr. C. D. Michener for the many 
manifestations of his support, encouragement and guidance throughout the course 
of this study. I am also especially indebted to Dr. G. W. Byers for his thoughtful 
consideration of the many problems and questions brought to him. 

I also offer very special thanks to Padre F. S. Pereira, CMF, who, besides 
suggesting this study in the first place, was my gracious host during a year's stay 
at the Departamento de Zoologia ("Museu Paulista") of the Secretaria da Agri- 
cultura in Sao Paulo, Brazil, and to Dr. Gonzalo Halffter, who on several occa- 
sions has in many ways facilitated my study of Mexican phanaeines. It is largely 
because of the generous offer of time, material and encouragement by these dis- 
tinguished students of Scarabaeinae that this study was possible. My sincerest 
thanks also to Dr. Paulo Emilio Vanzolini, Director of the Departamento de 
Zoologia, who in many ways facilitated my research activities in Brazil. 

My thanks also go to the following individuals who generously loaned material 
for this study from the collections in their care: O. L. Cartwright (United States 
National Museum), C. L. Hogue (Los Angeles County Museum), H. B. Leech 
(California Academy of Sciences), J. F. Lawrence (Museum of Comparative 
Zoology, Harvard University) and Roger Damoiseau (Institut Royal des Sciences 
Naturelles de Belgique). Dr. Damoiseau provided one of the two known speci- 
mens of Homalotarsus impressus for use in this study. 

Support of research activities in Brazil was provided by a Fulbright Scholar- 
ship (1965-66) awarded by the United States Department of State. Support of the 
final year of this study at the University of Kansas was provided by a Systematics 
and Evolutionary Biology Traineeship funded by the National Science Foundation 
and administered by the University of Kansas (Grant GB-4446X, Dr. William A. 
Clemens, principal investigator). The support of these agencies is most gratefully 
acknowledged. 

Finally, a special vote of appreciation to Mrs. Judy Keller, who typed the final 
version of the manuscript. 

PART II: SKELETAL MORPHOLOGY OF PHANAEUS V INDEX 

MACLEAY 

This a robust, compact dung beetle common in the eastern United States. 
The compactness of the body is accentuated by the strong sclerotization of 
nearly all external surfaces. Compared with related species, P. vindex is 
moderate in size, varying from 14 to 22 mm in length and 8 to 14 mm in 
width (at the bases of the elytra). Unlike most other scarabaeines, phanaeines 
are commonly brilliantly metallic in color. In this species most of the upper 
surface of the head and sides of the pronotum are yellow-green, the remainder 
of the pronotum is coppery-red, and the elytra, pygidium, exposed pteropleura 
and greater part of the lower body surface are green to yellow-green. 

The sculpturing of the integument varies in nature and degree over the 



738 The University Science Bulletin 

body surface. In general, the upper surface is strongly rugose; the lower 
surface is smooth to punctate but never rugose. Vestiture is restricted to 
pleural and sternal surfaces and to the appendages. Its variation is described 
below when appropriate. 

The Head 
head capsule 

As in other Scarabaeidae, the head is prognathous; the ventral and lateral 
walls are expanded, bringing the mouthparts into a horizontal plane with 
their apices directed forward. The capsule is very heavily sclerotized; as is 
common in beetles, the thickness of the integument tends to obscure many 
of the surface features (sutures, etc.) observed in other insects. 

The most conspicuous feature of the upper surface of the head of large 
(so called "major") males is a long, massive tapering projection, the cephalic 
process (Fig. 1, cpr), which arises between the eyes and curves posteriorly 
over the pronotum. The development of the cephalic process (as well as 
pronotal processes) is allometric and varies considerably within local popu- 
lations. In smaller ("minor") males, the cephalic process is considerably 
reduced in length (Fig. 2) or may be nearly absent. The cephalic process of 
the female, on the other hand, is less variable in development, being a short, 
erect, apically emarginate projection (Fig. 3). Large, often bizarre corniform 
outgrowths of the head wall (and pronotum) are very common in the 
Scarabaeidae, particularly in the subfamilies Scarabaeinae and Dynastinae 
(see Arrow, 1951). 

The flattened, anterior portion of the head dorsum is divided in front of 
each eye by an oblique lateral clypeal carina (Figs. 1, 4, lcrcly) extending 
from the head margin toward the base of the cephalic process. These carinae 
are almost always present in phanaeines but seldom in other dung beetles. 

Directly posterior and more or less parallel to each lateral clypeal carina 
is a faint line interpreted here as the fronto-clypeal sulcus (- f ronto-genal 
sulcus of authors) (Figs. 1, 4, fcs). Mesally this sulcus curves toward the eye 
and divides, one short, simple branch continuing toward but not reaching 
the eye, and the other, Y-shaped mesal branch of the fronto-clypeal sulcus 
(mbfcs) extending toward the base of the cephalic process. Medially the 
fronto-clypeal sulcus is incomplete. Ventral segments of the fronto-clypeal 
sulcus (Fig. 9) follow the anterior margins of the antennal sockets and 
terminate at the dorsal articulations of the mandibles (darm). A medially 
incomplete fronto-clypeal sulcus with a mesal branch is found in all 
Scarabaeinae. 

The fronto-clypeal sulcus marks an internal fronto-clypeal ridge (Figs. 6-7, 
fcr) joining the head wall above with that inflexed beneath it. Except via a 
small sinus, the transverse septum thus formed on each side of the head 



Evolution of the Phanaeine Dung Beetles 



739 




amcly 
clypr 




deye 



ocar 



Fics. 1-4. Phanaetis vindex. 1, head, lateral view, well developed, "major" male; 2, 
cephalic process of "minor" male, A — lateral view, B — frontal view; 3, cephalic process of 
female, A and B same as 2; 4, head of male, dorsal view (cephalic process removed). 



740 The University Science Bulletin 

prevents direct flow of hemolymph between the clypeus and paraocular areas 
of the parietals. The mesal end of each fronto-clypeal ridge forms a dorso- 
ventral brace between the upper head wall and the dorsal articulation of the 
mandible. Even though it is not an independent structure, for descriptive 
purposes this mesal bracing portion of each ridge will be called a cephalic 
brachium (head brace). Figure 5 shows the cephalic brachium (cbr). Be- 
cause its thickened vertical margins resemble two arms extending dorsally 
from the dorsal articulation of the mandible, the brachium resembles an in- 
verted bipod. Between its "arms" extends a vertical, triangular plate bent 
along its midline (height) such that a V-shaped line is inscribed where it is 
fused with the dorsal head wall. This line of fusion follows the arms of the 
Y-shaped mesal branch of the fronto-clypeal sulcus. The "bipodal" cephalic 
brachium and almost always coincident Y- or T-shaped mesal branch of the 
fronto-clypeal sulcus are characteristic of Phanaeus and Oxysternon. Other 
phanaeines and non-phanaeines examined have a "monopodal" brachium 
(see part III, character 7). 

Arising directly behind each dorsal articulation of the mandible is a 
slender, spatulate anterior tentorial arm (Fig. 5, ata). Originating on the 
expanded distal end of this process is a muscle inserting on the base of the 
scape. No anterior tentorial pits are visible. The anterior tentorial arms are 
very delicate and highly susceptible to loss if dissection is not careful. Very 
similar anterior tentorial arms were observed in all beetles examined. 

The clypeus is greatly expanded and, because of the ventral inflection of 
its margin, consists of two distinct areas: the flattened dorsal surface of the 
clypeus (Fig. 4, cly), anterior to the fronto-clypeal sulcus above, and the 
ventral surface of the clypeus (Fig. 9), anterior to the ventral portions of 
the same sulcus. Literature reference to the "clypeus" of dung beetles is 
usually to the dorsal surface. The dorsal surface of the clypeus of phanaeines 
is transversely ridged except for a smooth, narrow anterior border (Fig. 79, 
cly) ; the ventral surface bears setigerous punctures. 

The fold of the clypeus, separating dorsal and ventral surfaces and de- 
scribing the generally semicircular leading margin of the head is here referred 
to as the anterior margin of the clypeus (Figs. 4, 9, amcly). This margin is 
slightly upturned and, in vindex, slightly truncate or rounded medially. The 
morphological anterior limit of the clypeus, or apical margin of the clypeus 
(Fig. 9, apmcly), is the U-shaped margin of the ventral surface of the clypeus 
exposed when the labrum is removed. The dorsal articulations of the man- 
dibles (Figs. 5, 9, darm) are spherical condyles at the postero-lateral ex- 
tremities of the apical margin. 

The ventral surface of the clypeus is slightly depressed medially as a dis- 
tinctly crescent-shaped area which partially receives the labrum. Between 
this depression and the anterior margin of the clypeus is the clypeal process 



Evolution of the Phanaeine Dung Beetles 



741 



ata 




cbr 



apmcly 



arm 



deye 



cbr 



darm 





deye 



ocar 



ocfor 



Figs. 5-8. Phanaeas rindex, male. 5, eye region of head capsule, internal view; 6, trans- 
verse section through right side of head capsule anterior to fronto-clypeal sulcus (fs=plane of 
frontal section in Fig. 7); 7, oblique frontal section of fronto-clypeal ridge; 8, head capsule, 
caudal view (cephalic process removed). 



(= "estructura ventral del clipeo" of HalfTter, 1958 and elsewhere) (Figs. 9- 
10, clypr). The clypeal process of most dung beetles examined is, as in 
vindex, a rather strongly produced transverse carina. 

Interpretation of the surface areas of the head immediately posterior to the 
fronto-clypeal sulcus is difficult. The spatial relationship between the anten- 
nal sockets and this sulcus, however, suggests that a distinct frontal region 
(frons) has been much reduced or eliminated by anterior and/or posterior 
expansion of other cephalic areas (e.g., the clypeus). Consequently, a vertex 



742 The University Science Bulletin 

and genae are not distinguishable. These ideas are those of DuPorte (1960); 
it seems logical to adopt his designation of the dorsal and lateral areas of the 
head between the fronto-clypeal sulcus and the occipital area as the parietal 
area, or parietals. 

The expanded areas of the parietals posterior to the fronto-clypeal sulcus 
and lateral to the eyes are the paraocular areas of the parietals (Fig. 1, 4, 9, 
paar). Like the clypeus, they are inflected ventrally. Apparently because of 
their position, these areas are commonly referred to as "genae," a name with 
doubtful morphological connotations as applied to dung beetles. The para- 
ocular areas of most dung beetles are not "areas," but, rather, each is a finger- 
like extension of the head wall onto the eye surface. Such a prolongation is 
referred to as a canthus. 

In the phanaeines and a few other Scarabaeinae (e.g., Onitis and 
Scarabaeus), the paraocular areas of the parietals are expanded through the 
eyes, dividing them completely. Posteriorly each area firmly contacts, but is 
not fused with, a swelling of the parietals behind the eye, the postocular lobe 
of the parietals. The line of contact between these two expansions is a vertical 
"suture" connecting the posterior angles of both portions of the eye; this 
suture is the postocular line (Figs. 1, 5, pol). Figure 5 shows the internal 
structure of the head capsule of vindex, in this region. Here it can be seen 
that the internal margin of the paraocular area completely divides the optic 
foramen into upper and lower parts (oford, oforv). This division in vindex, 
and presumably in other dung beetles with similar structure, is correlated 
with a division of each optic lobe of the brain into two branches, each passing 
through a separate foramen. 

The shape of the posterior margin of the paraocular areas varies among 
phanaeines. As seen from above in vindex (Fig. 4), this margin curves to the 
posterior angle of the eye. In other phanaeines (e.g., Diabroctis mimas, Fig. 
115) it extends to the middle of the lateral margin of the eye, leaving, as seen 
from the side, a vertical flattened area (Fig. 115A, ca) completing the division 
of the eye. This flattened area will be referred to later as the canthal area. It 
suffices to note here that vindex and similar species do not possess a distinct 
canthal area. 

As seen from above (Fig. 4), strong depressions in the dorsal head wall 
behind the eyes set off the postocular lobes of the parietals (polb). These 
prominent lobes fit into notches in the anterior pronotal margin. Except for 
the dorsal portions of the postocular lobes which protrude from these notches, 
the dorsal surface of the head between the postocular lobes and posterior to 
the cephalic process is concealed by the anterior pronotal margin upon re- 
traction of the head. The cephalic-prothoracic interlocking mechanism 
achieved by accentuation of the postocular lobes of the parietals and notching 
of the anterior pronotal margin is characteristic of Phanaeus. Although it is 



Evolution of the Phanaeine Dung Beetles 743 

approached in Oxystemon, this mechanism is apparently not found in other 
phanaeines or other scarabaeines. 

The cephalic process of both sexes is a hollow outgrowth of the head wall. 
Because the fronto-clypeal sulcus is incomplete medially, it is impossible to 
know the exact morphological identity of the area of the head bearing the 
cephalic process. That is, it cannot be known with reasonable certainty 
where the clypeus ends and the parietals begin in this area. 

In some phanaeines, including vindex, there appears dorsally and laterally 
on the caudal portion of the head capsule a strong, bowed setose ridge (Figs. 
4, 8, ocr). It marks no corresponding internal ridge and probably represents 
a secondary thickening rather than a true occipital sulcus. Primarily because 
of its position, I call it the occipital ridge and the dorsal and lateral areas of 
the head wall between it and the postoccipital suture the occipital area 
(ocar). Similar occipital ridges occur in some other dung beetles; Matthews 
(1961) first referred to them as transverse occipital carinae in Copris. In 
American Canthonina the occipital ridge provides important characters at 
the generic level (HalfTter, in litt.). The morphological implications of "occip- 
ital" designations are justifiable since the occipital area is presumably, at 
least in part, composed of true occiput. Dorsally the occipital area is nearly 
suppressed because of the close approximation of the occipital ridge and post- 
occipital suture; laterally it is expanded, flattened and slightly depressed. 

The postoccipital suture (Fig. 8, pocs) is visible only dorsally as a fine 
line closely following the posterior margin of the head. Laterally it is ob- 
literated where it apparently coincides with the lateral margins of the occipital 
foramen. Thus, a very narrow postocciput appears only dorsally. Dorsally 
the postoccipital suture marks a deep, plate-like invagination continuous 
laterally with much narrower, rim-like inflections of the posterior margin of 
the head wall. This continuous inflection is the postoccipital ridge (Fig. 8, 
pocr). The ventral extremities of the postoccipital ridge are continuous with 
the tentorial structure described below. 

The occipital foramen is a dorso-ventrally elongate opening in the pos- 
terior head wall (Fig. 8, ocfor). Its ventro-lateral angles are rounded, and in 
them articulate the apices of the cervical sclerites. 

The ventral surface of the anterior portion of the head wall formed by 
the clypeus and paraocular areas of the parietals has been mentioned above. 
The antennal sockets are located in depressions of the paraocular areas 
marked posteriorly by the preocular carinae (Fig. 9, proc). A laterally- 
directed, conical antennifer (anf) projects from the mesal margin of each 
antennal socket. Antero-mesally the ventral portion of each eye closely ap- 
proaches the base of the mandible, from which it is separated by a narrow, 
vertical portion of the head wall, the oculo-malar space (= pleurostoma?). 
In some dung beetles (e.g., Scarabaeits) this area is separated by a weak 



744 



The University Science Bulletin 



armx 




ocar 



Figs. 9-15. Phanaetis vindex, male. 9, head, ventral view; 10, median portion of anterior 
margin of clypeus, frontal view; 11, ventral head wall, internal view; 12, dorsal process of 
tentorium, lateral view; 13-15, maxillary muscles originating on tentorium, diagrammatic: 
13, longitudinal (parasagittal) section (xs— plane of cross section in Fig. 14); 14, cross section; 
15, dorsal view (cf.. Fig. 11; ls=plane of longitudinal section in Fig. 13). 



suture extending from near the dorsal articulation of the mandible (i.e., the 
supposed position of the anterior tentorial pit) to the eye. This suture, which 
is obscured in phanaeines, is possibly part of a subgenal (pleurostomal) 
suture. 

Ventrally the posterior portion of the head capsule of scarabaeids is 
strongly convex. Its most conspicuous element is the gula (Fig. 9, gu), a 
large, central, bulging sclerite. The gula is located between the ventral ex- 
tensions of the postoccipital suture, which for convenience are referred to as 



Evolution of the Phanaeine Dung Beetles 745 

gular sutures (gus) (a cervical origin of the gula is assumed). The anterior 
external limit of the gula is approximated by a dense, transverse row of setae 
comprising the labio-gular fimbria (Fig. 9, nm). Removal of these setae from 
dung beetles reveals no suture, ridge or other line marking the junction be- 
tween the gula and submentum; it reveals only a line of change in sculpture 
between the smooth gula and setigerously punctate submentum. In other 
Scarabaeidae (e.g., Phyllophaga), however, an entirely similar, presumably 
homologous row of setae occurs along a definite labio-gular suture. It appears 
reasonable, then, to consider the labio-gular fimbria of Scarabaeinae the 
definitive line of fusion of the gula and submentum. The posterior margin 
of the gula (Fig. 8, imgu) is strongly inflexed and consolidated with the 
ventral extremities of the postoccipital ridge. The submentum is fused later- 
ally with the oculo-gular spaces and forms an integral part of the ventral 
head wall. 

The remaining portion of the head capsule mesal and posterior to the 
ventral portions of the eyes can best be considered as three indistinct paired 
areas (Fig. 9). The first two are the ventral extensions of the occipital area 
(ocar) and postocular lobes (polb). The third is the oculo-gular space (ogs), 
that area between the margin of the eye and the lateral margins of the gula 
and submentum. Because of the lack of ventral extensions of the occipital 
sulcus, the morphological composition of the oculo-gular space is indetermin- 
able; this space could be fused parts of the subgena or facial area (sensu 
DuPorte, 1960). If DuPorte is correct, the extent of the areas occupied by the 
ventral portions of the eyes of such scarabaeids as Phyllophaga indicates that 
the oculo-gular spaces are formed mostly by an invasion of the parietals onto 
the ventral surface of the head at the expense of the postgenae, which are 
greatly or completely replaced. This suggestion is further supported by the 
absence of an occipital ridge in many phanaeines and other dung beetles so 
that occipital and postgenal areas appear so consolidated with the parietals as 
to lose their identity. 

The antero-mesal margin of each oculo-gular space bears a solidly fused, 
dorsally flexed sclerite adjacent to the anterolateral angle of the submentum; 
this raised area is a hypostoma (Fig. 9, hyp). Dorsolaterally each hypostoma 
bears a small acetabulum, the ventral articulation of the 77iandible (Fig. 11, 
varm), which receives the mandibular condyle. 

The posterior tentorial pits (Fig. 9, ptd) are represented by wide depres- 
sions in the ventral wall of the head at the sides of the gula. Internal exami- 
nation of the gula, however, reveals a well developed tentorium supported 
laterally by invaginations along the entire lengths of the gular sutures rather 
than along the restricted areas indicated by the depressions. As DuPorte 
(1960) concludes for Phyllophaga, the gular sutures of dung beetles are 
probably best interpreted morphologically as elongations of the posterior 



746 The University Science Bulletin 

tentorial pits, and the gular ridges as expanded bases of the posterior tentorial 
arms. For descriptive clarity, however, use of the terms gular ridge and gular 
suture will be continued. 

The tentorium of vindex conforms to the very uniform structure found 
in the Scarabaeinae studied. The invaginations of the gular sutures, the gular 
ridges (Figs. 11, 14, gur) fuse medially to form a thinly sclerotized roof over 
the gula.'^ A strong antero-dorsal flexion of the posterior portion of this roof 
has produced what resembles a dorsally projecting platform supported 
medially by a vertical, longitudinal lamella. This platform is the dorsal 
process of the tentorium (Fig. 12, dprt). Posteriorly it extends as a very 
delicate, E-shaped posterior tentorial bridge (ptbr). Extremely delicate fila- 
ments connect the antero-lateral angles of the dorsal process with the anterior 
tentorial arms; in Figures 5 and 11 only incomplete portions of these fila- 
ments are shown. Originating on the dorsal process of the tentorium of 
vindex, and presumably other dung beetles, are two paired muscles. One 
(Figs. 13-15, c) inserts on the tendon of a much larger muscle (d) origi- 
nating posteriorly on the gula and inserting on the ventral posterior angle of 
the stipes; the other (b) divides apically into two sub-bundles, both of which 
insert on the cardo mesal to the maxillary articulation (Fig. 23). A third pair 
of maxillary muscles (a) originates posteriorly on the gular ridges and inserts 
on the articular process of the cardo lateral to the maxillary articulation. 
Muscles b, c and d are adductors of the maxillae; a is an abductor. 

The anterior margin of the tentorium (Fig. 11, amt) is thickened and 
resembles a bar arching between the internal antero-lateral angles of the gula. 
Associated with the ends of this "bar" are two mesally-directed processes, the 
maxillary articulations (armx), on which articulate the cardines. 

The hypopharynx forms with the prementum a compound structure 
treated below along with the labium. 

APPENDAGES OF THE HEAD 

As is characteristic of Scarabaeinae, the lamellate antennae are nine-seg- 
mented (Fig. 16-17). The posterior margin of the first segment, or scape 
(sp), bears a row of long setae, the scapal fringe (spfr), which reposes over 
the ventral surface of the eye. The second antennal segment is the pedicel 
(ped); the remaining seven segments comprise the Hagellum (fla). 

The last three antennal segments, the lamellae, together constitute the 
antennal club (Fig. 17). The form of the lamellae and antennal club is very 
characteristic of phanaeines and may be described as the Phanaeus-type. The 
basal lamella (seventh antennal segment) is scoop-shaped because of the 
concave distal surface which receives the smaller, similarly shaped second 
lamella. The terminal lamella (ninth segment) is received by the second; it 

3 In Anomiopsoides, a curious South American cucraniine, the gular ridges are not fused 
medially but are widely separated as in Phyllophaga. 



Evolution of the Phanaeine Dung Beetles 747 

is rounded and only slightly hollowed distally. Unlike the rest of the antenna, 
the lamellae are soft and fleshy except for the proximal, convex surface of the 
first, which is lightly sclerotized. 

Although such modification of the antennal club is characteristic of 
phanaeines, it is not restricted to them. An instance of convergence is the 
Phanaeus-typc antennal club of the South African genus Seepages. The an- 
tennal club of the male of the Bornean Onthophagus egregius Arrow is 
similar to the Phanaeus-type, while that of the Palaearctic geotrupine genus 
Lethrus is modified in a similar, though distinct manner. 

While the following account deals specifically with P. vindex, its mouth- 
parts are illustrative of peculiar, though quite uniform structures character- 
istic of the whole subfamily. The mouthparts of adult Scarabaeinae (and 
some Aphodiinae) are quite unlike their usually more generalized counter- 
parts in other beetles. As pointed out by Halffter and Matthews (1966) and 
others, they are highly modified for the manipulation of soft food, such as 
dung. The earliest discussion of adult scarabaeine mouthparts (those of 
Dichotomius [as Copris] carolinus) is that by Smith (1892), which deals 
primarily with the mandible. An excellent early study by Hardenberg (1907) 
compared the mouthparts of D. carolinus (as Pinotus) with those of other 
dung beetles, including Phanaeus carnijex {-vindex, not sulcatus), and with 
those of many other Scarabaeidae. Other general discussions include Mac- 
Gillivray (1923; Pinotus), Mohr (1930; Canthon, Bolboceros and Aphodius), 
Madle (1934; Aphodius), Halffter (1961; Canthon) and Halffter and 
Matthews (1966; Phanaeus). 

The mouthparts are here considered to include the labrum (= labrum- 
epipharynx of authors). Such consideration is supported partly by the mor- 
phological postulate that the labrum of insects is appendicular in origin but 
primarily by the curious morphology of the labrum of dung beetles, which 
suggests that it functions more as a mouthpart than a "preoral flap." The 
labrum ("labro-epifaringe") of Canthon is discussed at length by Halffter 
(1961). 

Except for the suspensorium, the labrum is a dorso-ventrally flattened, 
squared flap lying in a horizontal plane on the lower surface of the head. 
Figure 18 shows diagrammatically its position relative to the head wall. It is 
joined by a narrow membrane (arl, representing the labro-clypeal sulcus) to 
the apical margin of the clypeus. The median portion of this articular line is 
seen dorsally on the extracted labrum as a transverse sclerotic thickening 
(Fig. 19A, arl). Anterior to this line is the dorsal surface of the labrum 
(dsula), 4 which in other beetles is usually visible externally. 



4 Reference to position in discussion of mouthparts will be made in the definitive (prog- 
nathous) sense rather than the morphological (hypognathous) one. Hence, the morphological 
anterior surface of the labrum is called the dorsal surface, etc. 



748 



The University Science Bulletin 




ampr 



-mim^ 




ampr 



topr 



pmpr 



Figs. 16-19. Phanaeus vindex, male. 16, left antenna, ventral view (last two segments not 
visible); 17, same, side view of antennal club (7, 8 and 9 indicate segment number); 18, 
labrum and portion of clypeus, longitudinal (parasagittal) section; 19A, extracted labrum, 
dorsal view (muscles shown unilaterally; ls=plane of longitudinal section in Fig. 18); 19B, 
same, ventral (oral) surface. 



Evolution of the Phanaeine Dung Beetles 749 

Posteriorly the labrum extends internally as a strongly sclerotized sup- 
porting structure, the labral suspensorium (Figs. 18-19A). Posterior exten- 
sions of the labral suspensorium form three processes, two lateral formal 
processes of the labral suspensorium (topr) and a postero-median process 
(pmpr) which from above is strongly bent to the right. Extending to the 
apex of the labrum is the antero-median process (ampr, = "vastago clavi- 
forme" of HalfTter, 1952), a slender projection lying between the dorsal and 
ventral labral walls and adherent to the latter (Fig. 18). Apically the 
antero-median process expands dorso-ventrally, joining the upper and lower 
labral walls; ventrally this expansion bears the large setae of the median 
brush (mbr). Inserting laterally on each tormal process is a large muscle 
(lami) originating on the dorsal head wall directly behind the cephalic 
process; these muscles presumably depress the apex of the labrum. The 
insertion of these muscles suggests that the tormal processes represent the 
tormae (sensu Snodgrass, 1935) of other insects; indeed, the labral sus- 
pensorium might be the highly modified product of fusion of the tormae. 
A much shorter, second pair of muscles (lam2) extends between the anterior 
edge of a large round hole in the suspensorium and the membrane between 
the tormal and postero-median processes; they presumably pull this mem- 
brane forward. Band-like thickenings of the membrane posterior to the 
tormal processes connect the labral suspensorium laterally to the bases of the 
mandibles, maintaining a close appression of the labrum upon them. 

The lumen between the labral suspensorium and the ventral wall of the 
labrum (Fig. 18, lumen) contains an extensive network of exocrine glandular 
material, the secretions of which may serve such ends as brood ball con- 
struction (as an adhesive) or brood ball protection (as an antibiotic prevent- 
ing bacterial and fungal growth, etc.). 

The ventral surface of the labrum (Fig. 19B) is membranous and bears 
several types of regularly distributed setae. The median brush (mbr) is a 
longitudinal clump of thick, closely set setae inserted in the antero-median 
process of the labral suspensorium. Lateral to the median brush are similar 
setae scattered in a thick mat of short, fine hair. Two lateral, apically di- 
vergent longitudinal rows of prostrate setae comprise the lateral files (If). 
The space between the lateral files is the median area of the labrum ("zona 
central" of HalfTter, 1961); those spaces outside them, the lateral areas 
("zonas laterales"). The lateral areas are clothed by long, silky hairs similar 
to the longer ones along the apical margin comprising the apical fringe (af r) . 

The membrane extending posteriorly from the labral suspensorium (con- 
tinuous with the ventral surface) covers the upper ends of the molae of the 
mandibles and extends farther backward to form the dorsal wall of the 
buccal cavity. The ventral surface of the labrum rests against the mandibles 
with its median brush between the flattened incisor lobes of the mandibles. 



750 



The University Science Bulletin 



molb 



varsgal 




miac 



prmx 



Figs. 20-24. Phanaens vindex, male. 20, left mandible, lateral view; 21, same, ventral 
view, 22, molar lobes of mandibles, caudal view; 23, left maxilla, ventral view; 24, same, 
dorsal view. 



Of the mouthparts, the mandibles are perhaps the most curiously modi- 
fied (Figs. 20-21). Smith (1892), followed by Hardenberg (1907) and others, 
recognized several more-or-less distinct parts which he considered homologs 
of maxillary sclerites. For descriptive purposes, it is convenient to recognize 



Evolution of the Phanaeine Dung Beetles 751 

four mandibular parts: the basalis (Smith), incisor lobe, conjunctivus 
(Smith) and molar lobe. 

The basalis, or basal piece, is heavily sclerotized and bears the articu- 
lations and muscle insertions of the mandible (Fig. 20, ba). The mandibular 
acetabulum (mac, dorsal articulation) fits over the condyle on the postero- 
lateral extremity of the apical margin of the clypeus. The mandibular 
condyle (Figs. 20-21, mcon; ventral articulation) fits into the acetabulum in 
the dorso-lateral angle of the hypostoma. Directly lateral to the mandibular 
condyle is a small swelling on which is inserted the fan-shaped apodeme 
(apmni2) of the abductor muscle of the mandible; the larger apodeme of 
the adductor muscle (apmmi) is inserted ventro-mesally on a swelling 
adjacent to the molar lobe. Both of these muscles originate on the postero- 
ventral head wall. 

The receptacle of the mandibular glands (Figs. 20-21, rmgl) is thin- 
walled and cup-shaped; it is attached rather loosely to the posterior surface 
of the basalis. The receptacle contains a large mass of exocrine glandular 
tissue similar to that in the lumen of the labrum. Inserted medially on the 
receptacle is a large muscle (not shown in Figs. 20-21) with no conspicious 
apodeme which originates beside the adductor and abductor muscles on the 
head wall. Since its insertion is on a line between the dorsal and ventral 
mandibular articulations, it presumably does not function, at least efficiently, 
in gross mandibular movements. Rather, it likely moves the receptacle, 
perhaps to force discharge of the products of the mandibular glands. 

The molar lobes (Figs. 20-21, molb) are dorso-ventrally elongate, mesal 
lobes of the basalis. Their transversely ridged mesal surfaces are asym- 
metrical (Fig. 22), the left concave and the right convex. Miller (1961) 
found each molar lobe to consist of a series of juxtaposed microscopic 
processes ("tritors") capable of extreme trituration of food particles. Dorsally 
the ridging of both molar lobes becomes very coarse and projects beyond the 
posterior margin, which appears serrate (Fig. 20, morg). In situ the molar 
lobes fit against the vertical posterior wall of the hypopharynx, which forms 
the anterior portion of the deeply concave floor of the buccal cavity. Only 
the dorsal ends of the molar lobes project above the posterior hypopharyngeal 
wall. Presumably, food entering the buccal cavity is first masticated by the 
coarser dorsal ridging and then more finely ground by the microscopic ridg- 
ing after entering. 

The basalis extends anteriorly as a dorso-ventrally flattened lobe thought 
to function as a food scraper and to correspond to the incisor lobe of the 
more generalized chewing mandible. Laterally this incisor lobe of the 
mandible (inlb) is strongly sclerotized, but it becomes decreasingly sclero- 
tized and more translucent mesally. The apex of this lobe is repeatedly 
narrowly incised mesally and bears many long superficial setae (not shown 



752 The University Science Bulletin 

in Fig. 21). A double row of much shorter, closely-set setae comprising the 
comb of the incisor lobe (cmb; see Miller, 1961) occurs along the mesal 
margin of a weakly sclerotized, triangular area. This fringed area is referred 
to by Smith and several later authors as the "prostheca." 

The con]unctivus (Smith) is a convex, soft and flexible "sclerite" on the 
ventral surface of the mandible between the comb of the incisor lobe and 
molar lobe (Fig. 21, conj). It appears longitudinally ridged, but as Halffter 
(1961) has pointed out, each "ridge" consists of a row of thick adherent 
setae. In situ the conjunctivus fits over the thick, lateral margin of the 
posterior hypopharyngeal wall lateral to the molar lobe, possibly functioning 
as a seal or strainer preventing circumvention of the coarsely ridged dorsal 
ends of the molar lobes by large food particles. Miller (1961) believes the 
conjunctivus ("flexible area of the mandible") cushions the grinding move- 
ments of the molar lobes and, possibly, permits independent molar move- 
ments "while the mandibles are in the closed position." 

The maxillae articulate in a plane parallel to that of the mandibles. As 
seen ventrally (Fig. 23), they are well sclerotized appendages not markedly 
different from a generalized type, as are the labrum and mandibles. Each 
cardo (car) is rounded basally, and projecting mesally from its basal margin 
is the elongate articular process of the cardo (prcar). This process bears a 
forked lobe (prmx) which articulates with the maxillary articulation (Fig. 
11, armx). 

The stipes constitutes the bulk of the body of the maxilla. It is quadri- 
partite, each division being separated from the others by membranous sulci 
(i.e., the parts can be disjoined quite easily). What are here considered 
three distal sclerites of the stipes have been considered two subdivisions of 
the galea and a palpifer in other studies of scarabaeine morphology (e.g., 
Mohr, 1930). The primary basis for the present interpretation is the in- 
trinsic musculature of the maxilla, pertinent aspects of which will be brought 
out below. Although appropriate names for the parts of the stipes could be 
devised using descriptive prefixes, I have used numerical designations to 
avoid confusion with other maxillary terminology. 

The basal sclerite of the stipes, stipital sclerite I (Fig. 23, ssi; = stipes of 
authors) articulates with the cardo. On its internal surface originates a 
muscle inserted on the ventral articular sclerite of the galea (varsgal). 
Stipital sclerite II (ssn; = subgalea of authors) is a large, rectangular piece. 
Distally it bears the galea, laterally the lacinia; its ventral surface bears a 
wide oblique ridge, mesal to which the surface is strongly concave. The con- 
cave area fits upon the apex of the labium and is hidden when the maxillae 
are closed. Stipital sclerite III (ssm; = proxagalea of authors) is an elongate 
narrow piece interposed between the apices of sclerites II and IV. Stipital 
sclerite IV (ssiv; = palpifer of authors) is also elongate and forms the lat- 



Evolution of the Phanaeine Dung Beetles 753 

eral portion of the maxilla. Only sclerite IV is well developed dorsally (Fig. 
24). Originating on its base are two muscles, one inserting on the basal seg- 
ment of the maxillary palpus, the other on the dorsal articular sclerite of the 
galea (darsgal). 

The lobate galea (gal; = distagalea of authors) is articulated with the 
stipes by two transverse sclerotized bands, the articular sclerites of the galea. 
Both sclerites are free, one in the dorsal, the other in the ventral membrane 
of the base of the galea. Except for the articular sclerites, the galea is a thick, 
membranous lobe which is dorso-ventrally flattened and flexed ventrally 
over the apex of the labium. Dorsally it bears a dense pad of short, erect 
setae (Fig. 24). 

The lacinia (lac) is an apically widened and truncate, flattened lobe. 
Since it is united laterally with stipital sclerite II and not intrinsically mus- 
culated, it is presumably not moved independently. Dorsally can be seen a 
sclerotized strip which basally bears a dense tuft of rather long setae. The 
apex of the lacinia bears a dorsal, densely setose pad like that on the adjacent 
surface of the galea. 

The maxillary palpus (mxpl) articulates in a large socket in the apex of 
stipital sclerite IV. It is composed of four cylindrical segments which are 
successively longer distally. The apex of the distal segment bears a small, 
supposedly sensory, membranous spot. 

The cardo and stipital sclerites I and IV bear numerous long, stiff setae 
(not shown in Figs. 23-24). The remainder of the appendage, including the 
palpus, is either bare or with small, inconspicuous setae. 

The four extrinsic muscles of the maxilla insert on conspicuous apodemes. 
Those three originating on the tentorium (and gula) have been described 
above. The letters designating apodemes in Figures 23 and 24 correspond to 
those designating muscles in Figures 13-15. The fourth (Fig. 24, mlac) 
originates on the head wall and inserts on the base of the lacinia. 

As is generally true of beetles (Dorsey, 1943), the labial musculature of 
scarabaeines is considerably reduced and modified, making interpretation 
of the labium difficult by standard morphological criteria. Extrinsic muscles 
are absent and intrinsic ones are reduced in number. Likewise, the gross 
structure of the labium is curiously modified and difficult to compare with 
the generalized plan. 

The submentum is generally identified as that postlabial sclerite on which 
originate the median muscles of the labium (Snodgrass, 1935; = submento- 
premental muscles of Matsuda, 1965). This pair of muscles is either absent 
or considerably modified functionally in dung beetles. My designation of 
the submentum (smt) in Figure 9 and elsewhere is based primarily upon its 
position relative to the posterior tentorial pits (gular sutures) and its pre- 
sumed correspondence with similar, definitely premental plates in other 



754 



The University Science Bulletin 







27 



pmts 





mlbhpx 
mshpx 




Ishpx 



29 



mrghpx 



Figs. 25-29. Phanaeus vindex, male. 25, mcntum and prementum, ventral view (setae 
shown unilaterally); 26, prementum-hypopharynx, oral surface (setae shown unilaterally; floor 
of buccal cavity not shown, cf., Fig. 30); 27, right glossa, mesal surface; 28, median groove of 
prementum (glossae torn away; only apex of median lobe of hypopharynx shown); 29, 
hypopharyngeal sclerites, dorsal view. 

beetles. Snodgrass (1935:294) called the corresponding sclerite in Phyllo- 
phaga (Melolonthinae) the mentum, the submentum being represented by 
an internal plate on which originate the median muscles of the labium. He 
did not suggest the possibility that this internal plate represents only a pos- 
terior apodemal inflection of the submentum, the remainder of the sclerite 
being the exposed area distal to the gula. Kadic (1902), followed by Harden- 
berg (1907) and Mohr (1930), referred to the present submentum as the pos- 
terior plate of the submentum, the present mentum being the anterior plate. 
As described above, the submentum forms a part of the ventral head wall; 
distally it projects freely between the bases of the maxillae. 

The mentum (Fig. 25, mt) articulates with the straight distal margin of 



Evolution of the Phanaeine Dung Beetles 755 

the submentum. It is emarginate anteriorly and densely clothed with long, 
stiff setae. Its postero-lateral angles are inflexed mesally; on them articulate 
the ventral extremities of the hypopharyngeal suspensorium (Fig. 33). 

The remaining labial structures are borne on the dorsal surface of the 
mentum and, except for the palpi, are hidden. The prementum is repre- 
sented by a pair of premental sclerites (Fig. 30, pmts), the wedge-shaped 
ventral angles (vapmts) of which articulate with the antero-lateral angles of 
the mentum. In many dung beetles, including some phanaeines, the pre- 
mental sclerites are membranous medially and only a C-shaped margin is 
sclerotized (Fig. 150). 

The three-segmented labial palpi (Fig. 25, lbpl) articulate on the ventral 
angles of the premental sclerites (Figs. 25, 30). The two proximal segments 
bear many long, stiff setae, while the distal segment bears a single, very 
small seta on the outer surface. As does the distal segment of the maxillary 
palpus, the distal segment of the labial palpus bears a small, presumably 
sensory, membranous apical spot. On the basal segment of each is inserted 
a single muscle originating on the posterior margin of the adjacent pre- 
mental sclerite. 

Projecting anteriorly from between the premental sclerites are two elon- 
gate, membranous lobes, the glossae (Fig. 26, gl). The use of the term glossa, 
however, is loose and not meant to imply either that paraglossae are present 
or that they have presumably been lost. The glossal structure of all dung 
beetles and other scarabs examined is problematic. 

In all dung beetles examined, except Anomiopsoides, the mesal surfaces 
of the glossae are covered by a thick mat of short setae (Fig. 27) like that on 
the galeae and laciniae. Projecting mesally from the lateral margin of each 
glossa is the glossal flap (Figs. 26-27, glf), a translucent membranous flap 
fringed with long hairs. Seen mesally (Fig. 27), the glossae of phanaeines 
and some other dung beetles are apically bibbed. The slit separating these 
lobes is usually obscured by setae. One could assume that this slit is an 
incomplete division (latarima of MacGillivray, 1923) between the glossa and 
paraglossa, which are fused basally. However, consideration of the glossae 
of such scarabaeines as Heliocopris, which are trilobed, and of Catharsius, 
which are not notched, considerably weakens such an assumption. 

The glossae are supported laterally by the glossal bars (Figs. 26, gib). Pos- 
teriorly these bars project internally (Fig. 32, apglb) and are closely connected 
but not fused with the antero-dorsal angles of the lateral labial sclerites 
(Figs. 26, 30, libs). 5 Inserting on the connection of each of these two pieces is 

5 In Heliocopris the position of the lateral labial sclerites is occupied by a heavily sclerotized 
posterior extension of the glossal bar. This condition suggests that where lateral labial 
sclerites occur, they represent disjunct portions of the glossal bars. The strength of this sug- 
gestion, however, rests on a knowledge of the condition of the glossal bar in a wider variety 
of scarabaeines than that used for this study. 



756 



The University Science Bulletin 



pors 



mrghpx 



mrghpx 



vapmts 




Figs. 30-33. Phanaeus vindex, male. 30, extracted labium-hypopharynx (without sub- 
mentum), lateral view (broken outline refers to Fig. 32); 31, posterior wall of hypopharynx; 

32, suspensorium of hypopharynx and oral arms, lateral view (broken outline refers to Fig. 30); 

33, same, dorsal view (hypopharynx and prementum removed). 



a muscle originating medially on the anterior margin of the submentum 
(Fig. 13, e). Since each muscle diverges from its origin, I assume that, as a 
pair, they function primarily to spread the glossae. They could, however, be 
interpreted either as glossal muscles, whose origins have shifted from the 
presumably primitive prelabial position to the postlabium (submentum), or 
as the median muscles of the labium, which functionally have become ab- 
ductors of the glossae rather than levators of the prementum. 

In certain phanaeines, including vindex and some other dung beetles, the 
glossae come together at least basally (Fig. 26). In others, their bases are 
separated to varying degrees (Fig. 149). If the glossae are spread apart, the 
median groove of the prementum (Fig. 28) is exposed. The raised sides of 
this groove bear long, mesally-directed setae. Its floor is weakly sclerotized 
and bears dark, generally scattered sensory spots (sen) which, on internal 
examination, resemble sensilla coeloconia. 

Posteriorly the glossae are continuous with the hypopharynx, and together 



Evolution of the Phanaeine Dung Beetles 757 

they form a large, quadrate lobe on the dorsal surface of the mentum (Fig. 
30). The membranous median lobe of the hypopharynx (Figs. 26, 28, mlbhpx) 
is above the apex of the median groove of the prementum and slightly cov- 
ered laterally by the bases of the glossae. In species with separated glossae, 
the median lobe is completely exposed (Fig. 149) and covered with fine hair. 
Directly beneath the dorsal surface of the hypopharynx are three hypo- 
pharyngeal sclerites. A Y-shaped lateral hypopharyngeal sclerite (Fig. 29, 
lshpx) is on each side of the median sclerite (mshpx). Figure 29 shows their 
positions relative to the median lobe. Inserted on a hook-shaped ventral 
process of the median sclerite is a single (unpaired) muscle originating 
medially on the mentum (Fig. 13, f).° It supposedly depresses the median 
lobe of the hypopharynx. The vertical posterior wall of the hypopharynx is 
concave on each side of the postero-median ridge of the hypopharynx (Fig. 
31, mrghpx). This ridge is bowed to the left to fit between the asymmetrical 
molar lobes of the mandibles, which fit against the hypopharynx. 

Associated with the hypopharynx is a system of sclerotized bars com- 
prising the suspensorium of the hypopharynx (Figs. 32-33). This supporting 
structure is well developed in dung beetles and consists of two lateral arms 
(la; = suspensoria of Snodgrass, 1935) joined by a thicker, transverse bar. 
The lateral arms also support the membranous floor of the buccal cavity; 
their ventral ends articulate with the mentum. Extending dorsally from the 
transverse bar is the dorsal arm of the suspensorium (da). It projects freely 
through the body of the hypopharynx to beneath the dorsal extremity of the 
postero-median ridge (Fig. 32) and supposedly prevents collapse of the 
otherwise soft, membranous hypopharynx. 

The oral arms (Figs. 32-33, oa) are connected end-to-end with the lateral 
arms by thickened membranes. Ventrally they are bent mesally beneath the 
mouth (Fig. 33, mth). The dorsal ends of the oral arms are greatly expanded 
muscle discs ("tendon caps") on which insert short, thick muscles originating 
on the dorsal head wall behind the cephalic process. Since the arms are 
linked with the lateral arms of the suspensorium, which in turn articulate 
with the mentum, their muscles presumably effect elevation of the mentum 
(and prementum). From a functional morphological point of view, such 
function of these muscles, which are generally retractors of the hypopharynx 
(Snodgrass, 1935), is logical since median muscles (elevators) of the pre- 
mentum are apparently absent (see above). 

The paraoral sclerites (Fig. 30, pors) are oval, sclerotized thickenings in 
the lateral walls of the buccal cavity. On each is inserted a muscle from the 
head wall which presumably dilates the cavity. 



"Dorsey (1943) describes a similar unpaired muscle ("median muscle of the labium") in 
Popillia japonica (Rutelinae), but he finds it in no other beetle examined (p. 32), including 
Macrodactylas (Melolonthinae) . 



758 The University Science Bulletin 

The Cervix 

The cervix is the membranous region extending from the margin of the 
occipital foramen to the inflexed anterior margin of the prothorax. It is 
fairly long and allows the head to be forcibly withdrawn beyond the pro- 
thorax, although the head is probably never protracted this far by the living 
beetle. The cervix allows free movement of the head, but movement is 
largely restricted to a vertical plane by the tight fit of the posterior portion 
of the head within the prothorax. 

Located latero-ventrally in each side of the cervix is a single elongate 
cervical sclerite (Fig. 34). It articulates basally with the inflexed antero- 
lateral angle of the propleuron and apically with the rounded ventro-lateral 
angle of the occipital foramen. The setose external surface of the sclerite is 
concave; the internal surface is ridged for muscle attachment. Upon re- 
traction of the head, the apices of the sclerites fold dorso-mesally against the 
recessed anterior prothoracic margin. Situated dorsally in the cervix near 
the posterior head margin is a pair of lightly sclerotized, setose pads. Halffter 
(1961) has referred to similar pads in Canthon as anterior cervical sclerites. 
In view of their dorsal position and the double articulation of the cervical 
sclerite, it seems doubtful that they represent anything more than thickenings 
of the cervical membrane. 

The Thorax 
the prothorax 

The large, box-like prothorax accommodates the strongly developed 
muscles effecting movement of the head and front legs, the principal digging 
organs. Its anterior margin (to which is connected the cervical membrane) 
is recessed. The sclerites contributing to the anterior margin are all strongly 
inflexed and shaped to receive the posterior part of the head. 

Like the head, the pronotum of dung beetles and other "horned" 
scarabaeids often bear outgrowths (protuberances, ridges, etc.) producing 
forms characteristic of species. As are those of the head, such outgrowths are 
allometric and usually only the males bear well developed "armature." The 
pronotum of a well developed male of vindex is shown in Figures 35 and 
36; that of the female in Figures 3>7 and 38. In less well developed ("minor") 
males, its angulate appearance is less well marked. Sexual differences as 
exhibited by the prothorax are pronotal; the propleurae and sterna of both 
sexes are respectively alike. 

In phanaeines, a distinct postero-median angle of the pronotum (Fig. 38, 
5) fits between the bases of the elytra and covers the mesonotum. In some 
species this angle bears two distinct impressions, the posterior pronotal fossae 
(Fig. 194, pfos). On each side of the pronotum of all dung beetles (except 
some Canthonina— Halffter in litt.) there is a shallow, usually smooth de- 
pression, or lateral pronotal fossa (Fig. 36, lfos), marking an internal tubercle 



Evolution of the Phanaeine Dung Beetles 



759 




Figs. 34-38. Phanaeus vindex. 34, right cervical sclerite, lateral view (post, indicates 
posterior end); 35, prothorax of well developed male, dorsal view; 36, same, lateral view; 37, 
prothorax of female, lateral view; 38, same, dorsal view (1-5, sections and angles of circumnotal 
ridge: 1 — anterior portion, 2 — antero-lateral portion, 3 — lateral portion, 4 — antero-lateral 
angle, 5 — postero-median angle). 



on which articulates the base of the procoxa (Fig. 40, acx). In both sexes of 
vindex the pronotum is very strongly rugose (Fig. 165). 

The circumnotal ridge (Figs. 35-36, pntm) is a sharp ridge separating the 
definitive pronotum from the rest of the prothorax. For descriptive con- 
venience, various sections and angles of this ridge are defined in Figure 38. 
The sinuous lateral portion (3) is characteristic of the majority of phanaeines. 
The transverse anterior portion (1) is widened and in most dung beetles is 
continuous with the antero-lateral portions (2). In Phanaeus, however, the 
anterior portion is interrupted on each side by a distinct, rounded emargina- 



760 The University Science Bulletin 

tion (Fig. 35, emarg) into which fits the prominent postocular lobe of the 
parietals. The morphological significance of the circumnotal ridge is unclear. 
It is present in all scarabaeids and separates a dorsal area of the prothorax 
(pronotum) which differs usually in integumental sculpturing and often in 
color from the ventral areas (pleura and sternum). 7 

A very large, round opening in the flattened posterior wall of the pro- 
thorax (Fig. 39) fits over a protruding rim formed by parts of the meso- 
thorax. The lateral portions of this wall are widest and shaped so as to fit 
snugly against the opposing areas of the mesopleura. 

Except postero-laterally, the propleura (Figs. 36, 39, ppl) are ventral areas 
of the prothorax. Each propleuron is setigerously punctured except for a 
smooth, raised area lateral to the procoxal cavity. As seen ventrally (Fig. 39), 
their mesal margins are the curved lateral margins of the coxal cavities. A 
postcoxal extension (postcoxale) of each pleuron is joined with the lateral 
angle of the sternellum (stn) by a double pin-and-socket mechanism shown 
in Figure 42. Each propleuron is fused anteriorly with the lateral margin of 
the basisternum along the pleurobasistemal suture (Fig. 39, pbs). The 
propleura of a few phanaeines (Dendropaemon) and some other dung 
beetles (e.g., Copris and Earysternus) are each divided by a transverse 
propleural carina (Figs. 179-180, arrows). This carina extends laterally from 
the coxa, separating a "proepisternum" and "proepimeron." Since there is 
never an internal ridge corresponding to it, the carina is probably best in- 
terpreted as a secondary integumental feature rather than as a primary 
division between pleurites. In many dung beetles, including some pha- 
naeines, the anterior portion of each propleuron ("proepisternum") is 
strongly concave or otherwise set off from the "proepimeron." 

Coincident with the expansion of the prothoracic box, the procoxae have 
become transversely oriented, cylindrical structures whose apices nearly touch 
each other mesally. The propleura have expanded ventro-mesally and cover 
the bases of the coxae, which articulate internally with the lateral pro- 
notal walls. Their transverse orientation has affected the structure of the 
prostemum, a portion of which is depressed and hidden by the apices of the 
coxae. Thus, only a portion of the prosternum is visible with the coxae in situ. 

The exposed portion of the basisternum (Fig. 39, bst) is anterior to the 
front coxae and between the pleurobasistemal sutures. Its anterior margin 
(Fig. 41, imbst) is reflexed posteriorly and shaped to receive the gula and 



7 Butt (1944) has not differentiated notal and pleural areas in the prothorax of Amphimallon 
(Melolonthinae), but, rather, considers all prothoracic areas except the sternum as pronotal (p. 
11). There seems to be some validity to his designation (which could be applied equally well 
to dung beetles) if, for example, one negates the supposedly "homodynamous" evolution of all 
thoracic segments, particularly their pleura. Nevertheless, mainly because of its established 
use in beetle taxonomy, I choose to maintain at least descriptive distinction between pronotum 
and propleura. 



Evolution of the Phanaeine Dung Beetles 



761 




cxd 



ppl 



apst 



imbst 




imstn 



Figs. 39-43. Phanaeus tindex, male. 39, prothorax, ventral view (right coxa in place; 
setation shown on beetle's right only; xs=plane of cross section in Fig. 40); 40, same, cross 
section through right side; 41, prosternum, internal view (right prosternal apophysis removed; 
setation shown on left side only); 42, interlocking mechanism of sternellum and propleuron, 
right side (broken outline indicates cavities in sternellum receiving pleural lobes); 43, sternellum 
and adjacent parts, posterior view (mem=intersegmental membrane). 



adjacent occipital areas of the head. The basisternal crest (Figs. 39-40, 
bstc) is a median, longitudinal ridge extending posteriorly between the apices 
of the coxae to the anterior angle of the sternellum. A tuberculate swelling 
on each side of this crest articulates in a groove (Fig. 60, groove) in the apex 
of the coxa. These tubercles are secondary sternal articulations of the front 
coxae, primary articulations (with the lateral margins of the sternum) being 
hidden by the extensive overlap of the coxae by the pleura. The exposed 
portion of the sternellum (Fig. 39, stn) is triangular, the apex directed be- 



762 The University Science Bulletin 

tween the coxae and fused with the basisternal crest. Its posterior margin 
(Figs. 41, 43, imstn) is reflexed anteriorly and forms a shelf fitting over the 
mesosternum. 

Removal of the front coxae reveals the hidden portion of the presternum, 
a smooth trough on each side of the basisternal crest in which rotates the 
apex of the coxa (Fig. 39, cxd). Projecting from the lateral margin of each 
trough is a prosternal apophysis, a petiolate structure composed of three 
vertical lamellae (Figs. 39, 41, apst). To the apophyses are attached muscles 
from the posterior head margin, from the procoxae and from phragma I. 
The positions of these apophyses indicate that the depressed portions of the 
prosternum are composed of roughly equal parts of basisternum and sternel- 
lum, although there is no visible line between the two. 

Located latero-ventrally in the pro-mesothoracic intersegmental mem- 
brane on each side of the sternellum is a very large, spindle-shaped thoracic 
spiracle (Fig. 43, tspr). There is only one pair of thoracic spiracles. Between 
them is a V-shaped, setose interspiracular pad (ip), which, along with the 
spiracles, is probably never exposed during normal activity. Ritcher (1969a, 
b) discusses the spiracles of vindex and other scarabaeids. 

THE PTEROTHORAX 

Restriction of flight power to the hind wings is correlated with a pro- 
portionately large metathorax, which contributes the bulk of the volume of 
the pterothoracic box. Except for the mesepimera, the mesothorax forms a 
constricted rim over which fits the prothorax. Except in Anomiopsoides, 
which lacks hind wings, the pterothoracic structures of all scarabaeines ex- 
amined are quite similar. 

As seen dorsally, in situ (Fig. 44), the mesonotum appears as a single, 
pentagonal sclerite between the articular processes of the elytra. Much of it, 
however, is inflected and hidden by the close approximation of adjacent 
sclerites. Figures 45 and 46 show these external aspects of the mesonotum. 
Interpretation and naming of the mesonotal areas are difficult. A "V-suture" 
(Snodgrass, 1935 and elsewhere) separating scutum and scutellum is either 
lacking or unrecognizable. Furthermore, internal ridges, while well de- 
veloped, are probably not indicative of basic notal divisions. Primarily be- 
cause standard taxonomic terminology is based on it, it seems best to follow 
Snodgrass' interpretation (1909 and elsewhere) of the pterothoracic nota. 
Accordingly, the dorsally exposed, pentagonal mesonotal area is the scutel- 
lum (Fig. 44-46, sct2). In a few dung beetles (e.g., Eurysternus, Megathopa 
and Onitis) and in many other scarabaeids, a posterior, triangular portion of 
the mesoscutellum (usually very small in dung beetles) is permanently ex- 
posed between the bases of the elytra and not covered by the pronotum. The 
mesoscutum (Fig. 45, scu) is divided by the scutellum and represented on 



Evolution of the Phanaeine Dung Beetles 



763 



set 2 



epis2 




H£g^ 



Figs. 44-49. Phanaeus vindex, male. 44, pterothorax, dorsal view (elytra and hind wings 
removed; setation shown on right only); 45, extracted mesonotum, lateral view; 46, same, 
posterior view; 47, extracted mesopleuron, internal view: 48, extracted metanotum, anterior 
view; 49, same (stippling indicates internal structures). 



764 The University Science Bulletin 

each side of it by a rounded inflection of the notum on which articulates the 
first elytral axillary sclerite. A longitudinal, mid-dorsal line on the scutellum 
marks a deep, vertical plate (Fig. 46, vl) fused anteriorly to phragma 1. This 
plate intersects a narrower, horizontal plate (hi) extending between the 
scutal areas and also fused to phragma 1. 

The posterior infection of the mesonotum (Figs. 45-46, nt,,i) is strongly 
downflexed and widely emarginate. The origin of this margin is not clear; 
it may be scutellar, a remnant of the mesopostnotum, or of another origin. 
Functionally, however, it appears to be a mechanism allowing firm contact 
of the pteronota. The rounded (lower) margin of the inflection is joined 
very closely to the margin of phragma 2 by the intersegmental membrane. 
Articulating at each side of the inflection is a vertical, triangular plate corre- 
sponding to that which Snodgrass (1909) refers to as sclerite q (Fig. 46, q) 
in other beetles. Snodgrass suggests that sclerites q might represent the re- 
mains of a mesopostnotum. Inserted on it is a large muscle originating on 
phragma 1. 

At each side of the mesonotum, in front of the articular processes of the 
elytra, is a rectangular mesobasalare (Figs. 44, 50, 52, bas2). Each is produced 
subintegumentally as an acute apodeme (Fig. 52, abas2) on which originate 
two muscles, one inserting on the mesocoxal margin, the other on the base 
of the trochanter. 

The metanotum more nearly conserves typical pterothoracic structure 
than does the mesonotum. It is irregularly convex and much wider than the 
mesonotum. Figure 44 shows its surface areas and sutures. Like the meso- 
scutum, the metascutum (scu3) is divided medially by the metascutellum 
(sct3), from which it is separated by the V-suture (v). An oblique suture 
(s) on each side of the scutum sets off a triangular, postero-lateral scutal area, 
the lateral angle of which is the posterior metanotal wing process (pwp). 
Originating on each side of the metascutum are two large muscles, both of 
which insert on the anterior margin of the metacoxa. The prescutum (Fig. 
48, pscu) forms an anterior surface of the metanotum hidden in situ by the 
mesothorax and separated from the scutum on each side by a weak, trans- 
verse w-suture (Figs. 44, 48, w). Lateral projections from the prescutum are 
the anterior metanotal wing processes (awp). The paraphragmal areas of 
the prescutum (Fig. 48, ppha) are fused with the lateral margins of phragma 
2 along the line of attachment of the intersegmental membrane, and the 
composite structure (i.e., paraphragmal areas + phragma 2) is separated 
from the body of the metanotum by a narrow, transverse membranous area 
(mem). For clarity, Figure 48 is modified in Figure 49 to contrast internal 
(apodemal) metanotal areas (stippled in Fig. 49) with external ones. In- 
serted on the paraphragmal areas are muscles originating on the metasternum 
and comprising the tergo-sternal indirect flight muscles. The metapostnotum 



Evolution of the Phanaeine Dung Beetles 



765 



epim 




st 3 



S1+2 



Figs. 50-51. Phanaeus vindex, male. 50, pterothorax, lateral view (elytra and coxae in 
place; mem=intersegmental membrane); 51, same, ventral view (right coxae removed). 



(Figs. 44, 53A, pstn) is reduced to a narrow ridge united with the posterior 
metanotal margin and bearing phragma 3. 

The mesopleura are compressed. The mesepisterna (Figs. 50, 52, epis2) 
form the sides of the projecting mesothoracic rim fitting within the pro- 
thorax; basally they are united with the mesosternum. A ridge -like inflection 
of the anterior margin of each mesepisternum (Fig. 47, imepis) is fused 
dorsally and ventrally with the mesopleural ridge. On the dorsal end of this 
inflection is a disc (md) on which originates a muscle inserting on the 
antero-lateral angle of the mesonotum. The mesopleural suture (Fig. 52, 
pls2) is an impressed, vertical groove separating the mesepisternum and 
mesepimeron. Its internal invagination is the mesopleural ridge (Fig. 47, 
plrg), the prominent dorsal and ventral ends of which are, respectively, the 
mesopleural wing process (wp2) and pleural articulation of the mesocoxa 
(acx 2). The pleural ridge is the origin of a large muscle inserting on the 
mesocoxal margin. Each mesepimeron has three distinct surfaces, anterior 



766 



The University Science Bulletin 




st3 



scu3 



sct3 




Figs. 52-53. Phanaeus vindex, male. 52, pterothorax, anterior view (elytra and mesocoxae 
in place; setation shown on right only); 53A, same, posterior view (elytra and metacoxae re- 
moved); 53B, dorso-posterior angle of metasternum, lateral view. 



(Fig. 52, epim2 a ), lateral (Fig. 50, epim2i) and posterior (Fig. 47, epimi,), 
which are set off by sharp, vertical bends in the body wall. Of the three, 
only the lateral surface is permanently exposed and colored and sculptured 
like adjacent exposed sclerites. The posterior surface is pressed close to the 
anterior surface, such that in Figure 47 care must be taken to note the 
vertical posterior margin of the mesepimeron (epim P m) to distinguish in- 
ternal and external surfaces. A smooth, elongate area (y) on the posterior 



Evolution of the Phanaeine Dung Beetles 767 

surface receives the opposing margin of the metepisternum; a similarly 
smoothed, concave area of the lateral surface (z) contacts the base of the 
mesocoxa. 

The metapleurites are oriented horizontally (see Snodgrass, 1909). The 
metepisternum has two distinct surfaces, anterior (Fig. 57, epis a ) and lateral 
(Figs. 50, 54, epis3i). The lateral surface is trapzoidal, and its dorso-posterior 
angle (Figs. 54, 56, tab) is produced as a rounded metepisternal tab which 
fits in a depression in the elytral epipleuron. The metepisternal tab is char- 
acteristically present in phanaeines. The metepimeron (Fig. 54, epim) is a 
narrow, horizontal sclerite along the dorsal edge of the episternum. It is 
completely hidden beneath the epipleuron of the closed elytron. Posteriorly, 
a finger-like portion of the epimeron (Figs. 53A, 56) projects mesally and 
unites with the lateral extremity of the metapostnotum. Separating the 
metapleurites is a faint horizontal line, the metapleural suture (Fig. 54, pis). 
Its invagination, the metapleural ridge (Fig. 55, plrg) is a very strongly 
developed, oblique plate; Figure 55B shows the origin of this ridge relative 
to the metapleurites. The antero-dorsal extremity of the metapleural ridge is 
the metapleural wing process (Figs. 54-57, wp), which supports the second 
axillary sclerite and base of the Sc vein of the hind wing. Posteriorly the 
metapleural ridge unites with a vertical ridge (Fig. 55A, rg) supporting the 
mesal projection of the metepimeron, and the two converge on a rounded 
lobe (Figs. 54-55, lobe) which inserts in a slot in the dorsal margin of the 
metasternum. This lobe is not the pleural articulation of the metacoxa. The 
functional "pleural" or dorsal articulation of the metacoxa (Figs. 53A-53B, 
dacx3) is a tubercle borne on the dorso-posterior angle of the metasternum. 
Whether or not this tubercle is the dissociated extremity of the pleural ridge 
or a secondary articular structure is not clear. 

The metabasalare (Fig. 57, bas) is fused with the anterior surface of the 
metepisternum, and together the two form an irregular vertical wall fitting 
against the posterior surface of the mesothorax. The stippled area in Figure 
58 (c.f., Fig. 57) is the internal portion of this anterior wall. (In some 
beetles [e.g., Carabidae] the metabasalare is free, but in Scarabaeidae it is 
apparently always fused to some degree with the metepisternum [see Snod- 
grass, 1909 j.) The apex of the basalare (Fig. 59, basp) is elongate and closely 
associated but not fused to the metapleural wing process (wp); it provides a 
second fulcral point for the base of vein Sc of the hind wing. Internally the 
base of the basalare bears an enormous, oval muscle disc (Figs. 56-57, basmd). 
From this disc extend two muscles, a very large one to the metasternum and a 
much smaller one to the posterior margin of the metacoxa. Since the basalare 
is rigidly fused with the body wall, it is quite likely that the basalar-meta- 
sternal muscles augment the action of the tergo-sternal indirect flight muscles. 

The mesosternum (Fig. 51, st2) is a narrow, transverse sclerite between 



768 



The University Science Bulletin 




Figs. 54-59. Phanaeus vindex, male. 54, right metapleuron, lateral view; 55A, same, in- 
ternal view (basalare removed; xs=p!ane of cross section in 55B); 55B, same, cross section; 
56, same, dorsal view (basal membrane of hind wing removed); 57, same, anterior view; 
58, same as 57 (stippled area is internal); 59, apices of metabasalare and metapleural wing 
process, inner view. 



the apices of the mesocoxae. It is convex and bears a transverse, brace 
( ) -shaped ridge which contacts the inflexed posterior margin of the 

prosternellum. The mesosternum is fused with the median area of the meta- 
sternum along the mesometasternal suture (sts). 

The very large metasternum (Fig. 51, st3) forms most of the ventral wall 
of the pterothorax. It is superficially divided into three parts by the elongate 
mesocoxae, which rotate in depressions of the metasternum (cxd) similar to 



Evolution of the Phanaeine Dung Beetles 769 

the procoxal depressions of the presternum. The shape of the median area 
of the metasternum (between the mesocoxae) is characteristic of phanaeines. 
Slightly posterior to the mesometasternal suture, it is strongly and abruptly 
depressed medially as a vertical angulate prominence. Consequently, as seen 
laterally (Fig. 50), the metasternum is flat and squared anteriorly; seen ven- 
trally (Fig. 51), it is pentagonal and acuminate anteriorly. Since this depres- 
sion of the median area is normally viewed from below, the metasternum of 
phanaeines is often described as "raised." 

THE PTEROTHORACIC ENDOSKELETON 

The pterothoracic structures of interest here are the phragmata and 
sternal apophyses; other internal skeletal features are discussed above. 

Phragma 1 (Figs. 45-46, 52, phi) is a deeply invaginated plate borne by 
the anterior margin of the mesonotum. Medially its ventral margin is bent 
posteriorly, resulting in the bilobed appearance seen anteriorly (Fig. 52). On 
its anterior surface originate a pair of muscles inserting on the prosternal 
apophyses; each side of the posterior surface is the origin of two muscles, a 
large one inserting on the mesocoxal margin and a smaller one on sclerite q. 
There are apparently no well developed muscles extending between phrag- 
mata 1 and 2. Phragma 2 (Fig. 48, ph2) was mentioned previously. 

Phragma 3 (Fig. 53A, ph3) is borne by the metapostnotum. Like 
phragma 1, it is emarginate medially for passage of the aorta and digestive 
tract. At the sides of the emargination are two dark, oblique lines which 
separate a median area; from this area extend short muscles (the dorsal 
longitudinal indirect flight muscles) to phragma 2. It is noteworthy that 
these muscles, which effect the downstroke of the wing, are so poorly de- 
veloped compared to the extensive development of their antagonists, par- 
ticularly the tergo-sternals and basalar-sternals. From the lateral areas of 
phragma 3 extend muscles, the "lateral obliques," inserting along the anterior 
margin of the metascutum. On the lateral processes of phragtiia 3 (ph3ip) 
attach muscles from the metafurcal arms that probably stabilize the meta- 
furca. 

Both the mesosternal and metasternal apophyses are fused basally. The 
resulting compound structures, the mesofurca and metafurca, are quite dif- 
ferent in shape. The base of the mesofurca (Fig. 52, bfur2) is a transverse, 
vertical plate marked externally by the mesometasternal suture. In phanaeines 
and some other scarabaeines, it is braced by a vertical, longitudinal plate, the 
mesof ureal brace (br). The mesof ureal arms (afur2) are cup-shaped, open- 
ing laterally; within them and on their posterior surfaces originate muscles 
inserting on the mesocoxal margin. A ribbon-like, lateral extension of each 
arm (afur2p) serves as the origin of a muscle extending through the coxa 
and inserting on the base of the trochanter. 



770 The University Science Bulletin 

The metafurca (= metendosternite of Crowson, 1938, 1944) is an an- 
teriorly inclined, T-shaped invagination of the metasternum (Fig. 53A). It 
is borne by a rounded, sternal inflection, the metafurcasternum, (fst3), be- 
tween the bases of the hind coxae. The tubercle on each side of the fur- 
casternum is a ventral (sternal) articulation of the metacoxa (vacx3). The 
base of the metafurca (bfur3) is elongate and, in cross section, T-shaped (as 
a steel T-beam). The metaf ureal arms (afur3) are subapical (the apex of 
the metafurca is not visible in Fig. 53). On the metafurca, as a whole, orig- 
inate five pairs of muscles, all of which insert on the metacoxa. 

THE LEGS 

The fossorial front leg of vindex is shown in Figures 60 and 61. The coxae 
(ex) are cylindrical and articulate laterally with the internal, tuberculate 
lateral articulation of the procoxa (Fig. 40, acx), which fits into a small 
notch (Figs. 60-61, n) in the coxal margin. The concealed trochantin (Fig. 
62) articulates anteriorly on the coxal margin. The trochanter (Figs. 60-61, 
tr) is fused to the base of the femur and largely hidden in the apex of the 
coxa; the femur-trochanter line of fusion is visible only posteriorly. The 
inner margin of the femur (fe) is flattened to allow close appression of the 
tibia. The profemoral hair plate (Fig. 60, fet) is a circular group of closely 
packed, presumably proprioceptive setae on the anterior surface of the femur. 
When the femur is retracted, this hair plate slips behind the apex of the coxa. 
The tibia (tib) is flattened antero-posteriorly and bears three lateral protibial 
teeth (tibt). The anterior (atibc) and posterior tibial carinae (ptibc) (Figs. 
60-61, respectively) extend the length of the tibia. Abrupt, lateral turns of 
the anterior carina extend onto the surface of the two distal tibial teeth. In 
the discussions that follow, tibial teeth bearing such extensions of the anterior 
tibial carina will be called carinate. A single protibial spur (tibs) articulates 
on the inner apical angle of the tibia; in vindex and most other phanaeines, 
it is pointed and only slightly curved. In other dung beetles, the spurs may 
show sexual differences (e.g., some Canthon) or extraordinary modifications 
(Onitini), but they are apparently rarely absent. 

In most dung beetles protarsi are well developed and found in both sexes. 
In phanaeines, however, they are found only in the females of certain groups, 
including Phanaeus, where they are always reduced in size and clawless 
(Fig. 63). When present, it inserts on the inner apical angle of the tibia, 
mesal to the tibial spur. No known male phanaeines possess front tarsi; their 
curious lack in phanaeines and some other dung beetles has been known for 
many years (Brulle, 1837; Dejean, 1838). 

The middle and hind legs differ principally in length and in articulation 
and orientation of the coxae. The middle leg (Fig. 65) is shorter and 
stockier than the hind leg (Fig. 66). The cylindrical mesocoxae (Figs. 51, 



Evolution of the Phanaeine Dung Beetles 



771 



groove 




Figs. 60-66. Phanaeus vindex, female. 60, left foreleg, anterior view; 61, same, posterior 
view; 62, trochantin, anterior view; 63, front tarsus; 64, right middle tarsus, ventral ( = 
plantar) surface; 65, right middle leg, ventral view; 66, right hind leg, ventral view (coxa 
removed; see Fig. 50). (Setae are shown only in Figs. 63-64.) 



cx2, and 65, ex) are oriented longitudinally and approximately parallel in 
most phanaeines. The metacoxae (Fig. 51, c\5) are elongate and always 
oriented transversely. The middle coxa has but a single, pleural articulation 
(to allow rotation along its long axis), while the hind coxa has two. Neither 
posterior pair of legs has a recognizable trochantin (contrary to Snodgrass, 



772 The University Science Bulletin 

1909: 563). The trochanter (tr) of both pairs is triangular and fused to the 
base of the femur. Their femora differ essentially only in length. The tibiae 
(tib) are only slightly shorter than their respective tarsi. They differ in 
shape, the middle tibia being round in cross section, the hind tibia being 
approximately rectangular because of its flattened dorsal surface. The mid- 
dle tibiae each bear two unequal apical tibial spurs (tibs) ; each hind tibia, 
only one. The lateral surface of the hind tibia of a few phanaeines bears a 
subapical transverse tibial carina (Fig. 262, cr), which is absent in vindex 
(Fig. 261). This carina is uncommon in phanaeines but is characteristic of 
certain other groups, such as Copris. Middle and hind tarsi are well de- 
veloped but always flattened and clawless in phanaeines. Figure 64 is a 
posterior tarsus of vindex; as in all phanaeines, the ventral (plantar) surface 
of each tarsomere except the last is longitudinally carinate, while the dorsal 
surface is smooth. 

THE WINGS AND THEIR ARTICULATIONS 

The elytra (Fig. 67; 50, ely) completely cover the metanotum, metepi- 
mera and all but the last abdominal terga. The humeral angle (h) of each 
elytron is swollen; beneath it lies the raised base of the folded hind wing. 
The elytral epipleitron (Fig. 50, epip) is the downflexed lateral margin of 
the elytron; in phanaeines it is narrow. The elytra of phanaeines are held 
tightly in place when at rest by two complementary mechanisms, namely, 
by the interlocking grooves in the opposing elytral margins along the elytral 
suture and by containment of each elytron in the epipleural shelf by the 
metepisternal tab. The epipleural shelf (Fig. 56, epips) is formed by the 
recessed dorsal margin of the metepisternum, on which rests the epipleuron. 
The anterior margin of each elytron is strongly downflexed (Fig. 52), form- 
ing a vertical wall fitting against the prothorax. The articular process of 
the elytron (Fig. 67, apely) is a lobate extension of the flexed anterior margin. 
The elytral striae (Fig. 67, st) are longitudinal, impressed lines. The elytra 
of phanaeines each have eight striae, including the often overlooked one (8) 
adjacent to the epipleuron. The striae are numbered as in Figure 67 (1-8). 
Each longitudinal zone between two striae is an inter stria (= interstice, in- 
terspace). The interstriae of vindex are flat and coarsely rugose (Fig. 284). 

There are two elytral axillary sclerites (axillaries) beneath the hollowed 
underside of the articular process of the elytron (Fig. 68). The first (Figs. 
68-69, axl) articulates with both the mesoscutum (at point a in Fig. 68) and 
the mesopleural wing process (along ridge b). This double articulation sug- 
gests that the first axillary is the functional equivalent of the first two axil- 
laries found in other scarabs (e.g., Phyllophaga; Snodgrass, 1909), which 
may be fused in phanaeines. The concave second axillary (sod) fits over the 
first and is connected by a membranous thickening with the small flexor 
sclerite of the elytron (Fig. 69, fs), which projects internally as a conical 



Evolution of the Phanaeine Dung Beetles 



773 




apely 



apely 




axi 



fs 



ax2 



>c; 



/\ 




69 



ax2 



axi 



ax3 




71 




j ug a I lobe 



vannal lobe 



Figs. 67-71. Phanaerts i index, male. 67, left elytron, dorsal view (strial number, 1-8, to 
left of corresponding stria); 68, articular process of elytron and associated elytral axillary 
sclerites, ventral view (a=lobe of axillary 1 articulating with mesoscutum; b=ridge of axillary 
1 articulating on mesopleural wing process); 69, extracted elytral axillary sclerites, dorsal view; 
70, base of right hind wing and adjacent metanotal area, dorsal view; 71, right hind wing, 
dorsal view (bc=basal cell: other symbols indicate veins). 



apodeme. Inserted proximally (nearest the body wall) on the flexor apodeme 
is a thin "tendon" (no muscle) originating on an apodeme (Fig. 57, pr) 
from the lower angle of the metabasalare. Inserting distally is a muscle- 
bearing, flexible apodeme originating on a process (Fig. 47, pr) from the 
mesepimeron. The insertions of the direct muscles of the elytron are difficult 
to trace, but all originate on the mesopleural ridge. 

The membranous hind wings are well developed; dung beetles are gen- 



774 The University Science Bulletin 

erally strong flyers. The reduced venation is of the scarabaeoid type discussed 
by Stellwaag (1914), Forbes (1922), Graham (1922), d'Orchymont (1920), 
Balthasar (1942), et al., no two of whom fully agree on vein terminology. 
The system used here (Fig. 71) is largely a compromise of several systems. 
The vein and area designations used are not meant to imply homology with 
those of other insects, including beetles; this terminology is adopted here 
strictly for comparative purposes within the group. 

There are three axillary sclerites of the hind wing (Fig. 70). The first 
(axl) hinges laterally on the anterior metanotal wing process and articulates 
anteriorly with the base of vein Sc. The small second axillary (ax2) articu- 
lates ventrally on the metapleural wing process. With the wing extended (as 
in Fig. 70) the apex of the third axillary (ax3) is directed upward, beyond 
the wing surface. When the wing is flexed, the third axillary rotates on its 
long axis, the apex swinging mesally to above the posterior metanotal wing 
process. Mesal to the third axillary and connected to it by a thickening in 
the basal wing membrane is the flexor sclerite of the hind wing (fs). On the 
flexor sclerite insert all three of the direct flight muscles of the hind wing, 
two of which originate on the metapleural ridge, the third on the dorsal side 
of the metabasalar muscle disc. 

As external sclerites, subalares are lacking in the mesothorax, unless 
represented by the flexor sclerites. In the metathorax, however, the subalare 
is definitely represented by a stalked muscle disc (Fig. 53A, sub3 m d) inserted 
on the basal membrane of the hind wing posterior to the third axillary 
(point i in Fig. 70). From this disc extends a muscle to the anterior margin 
of the metacoxa. 

The Abdomen 

the abdominal terga 

The dorsum of the abdomen is composed of eight terga (Fig. 72, Ti~- 
and pyg). The first seven are covered by the elytra and are thinly sclerotized 
except for the acrotergites (at), thin bands across the anterior margins of 
each. The ridge-like invagination of each acrotergite is an antecosta, an at- 
tachment for dorsal longitudinal muscles. The acrotergite of tergum 1 is the 
metapostnotum; its antecosta, phragma 3. Associated latero-ventrally with 
each of the first seven terga is a pair of abdominal spiracles (spri-7). With 
few exceptions, these spiracles of phanaeines are located in the lateral mem- 
branous region separating terga and sterna. 8 The large elliptical spiracles of 
tergum 1 (Fig. 73) open anteriorly; the much smaller succeeding six pairs 
open dorsally. 



8 This arrangement of abdominal spiracles is characteristic of "Scarabaeidae — Laparosticti," 
a commonly employed but informal taxonomic division of the family including Scarabaeinae 
(see Ritcher, 1969b). 



Evolution of the Phanaeine Dung Beetles 



775 




(post.) 

Figs. 72-78. Phanaeus vindex, male. 72, abdomen, dorsal view; 73, first abdominal spiracle, 
anterior view; 74, abdomen, caudal view; 75, same, lateral view (terga 1-7 removed); 76, 
genital capsule, lateral view; 77, same, dorsal view; 78, ninth sternum, ventral view (post, in- 
dicates posterior margin). 



The last (eighth) abdominal tergum, or pygidium, is a thickly sclero- 
tized, vertical plate covering the anal and genital openings 9 ; it is only par- 
tially covered by the elytra. The covered portion, the base of the pygidium 
(Figs. 72, 74-75, bpyg) is short and grooved medially. The permanently ex- 
posed apex of the pygidium (apyg) is rounded and, in phanaeines, bears a 
continuous, unbroken raised margin. 



9 The suggestion arises that the pygidium comprises two terga, 8 and 9; however, a critical 
study of the musculature of the pygidium and adjacent structures (which is quite intricate) is 
in order before the possibility can be seriously considered. 



776 The University Science Bulletin 

the abdominal sterna 

The abdominal venter is composed of eight sterna in the female, nine in 
the male. The first two (Figs. 51, 75, Si + i;) are inflexed behind the meta- 
coxae and, except for the dorso-lateral extremities of the second, permanently 
hidden. 10 Sterna 3-8 are ventrally exposed (Fig. 75, S.s-s). Although clearly 
separated by impressed lines, they are rigidly fused. Sterna 1-3 are acutely 
produced antero-medially between the apices of the metacoxae (Figs. 51, 75) 
and form a prominence fitting against the metafurcasternum. The dorso- 
lateral extremities of sterna 2-7 (= abdominal pleurites of authors) are im- 
pressed and covered by the elytra. Longitudinal ridges (Fig. 75, rg) separate 
these impressed areas from the ventral portions. The dorso-lateral ex- 
tremities of sternum 8 are acute and do not extend beneath the elytra. 

Sternum 9 (Fig. 78) is found only in the male. It is a thinly sclerotized, 
internal plate (i.e., "cryptosternite") fitting beneath the apex of the genital 
capsule. Though no such plate is present in the female, it may be repre- 
sented by a pair of poorly sclerotized plates below the opening of the 
common oviduct. 

EXTERNAL GENITALIA 

The male intromittant organ is enclosed in the genital capsule (Figs. 76- 
77), which consists of an elongate tube, the phallobase (phb), bearing api- 
cally a pair of lateral appendages, the parameres (pm). Within the phallo- 
base is a well developed, membranous lobe, the aedeagus, which is presum- 
ably extruded from the phallobase between the movable parameres during 
copulation. Basally the aedeagus bears a large aedeagal apodeme (ap), on 
which insert the retractor muscles of the aedeagus. Sclerotized external 
genitalic structures exist only in the male of dung beetles. The terminology 
applied here is that of Snodgrass (1957). It differs considerably from that 
commonly applied in beetle taxonomy, in which, for example, the genital 
capsule is often referred to as the aedeagus, and the present aedeagus as the 
median lobe, phallus or penis. Lindroth (1957) has reviewed the various 
terminologies applied to beetle genitalia. 

PART III: COMPARATIVE SKELETAL MORPHOLOGY 

OF PHANAEINES 

The purpose of this part is to present a resume of the morphological 
variation largely upon which is based the classification proposed in part V 
and the evolutionary discussion in part VI. In addition, this section intro- 



111 The second abdominal sternum, when exposed only dorso-lateral ly, characterizes the 
haplogastran type beetle abdomen; Jeannel and Paulian (1944) stressed this character when 
grouping Scarabaeoidea and Staphylinoidea as the suborder Haplogastra (but see Crowson, 
1955). 



Evolution of the Phanaeine Dung Beetles 777 

duces some characters which, to my knowledge, have heretofore not been 
discussed in systematic studies of Scarabaeinae. 

The characters treated here are admittedly only a small portion of those 
potentially available from an unlimited study of dung beetle skeletal mor- 
phology. Practicality, however, demands adoption of some reasonable 
criteria to limit them to a manageable number, which, unless computer 
techniques are used, is apt to be quite small. The following criteria were 
stressed in selecting characters discussed below: 

(1) that they (the characters) be supraspecific characters; that is, that 
their variation in general be less within a species group (or higher category) 
than among several. This criterion, however, does not strictly limit charac- 
ters to those with wide taxonomic application since some character states 
may be constant only among the species of a single taxon while varying 
widely among the species of other taxa of similar size or rank; 

(2) that as nearly as possible they represent all tagmata and appendages; 

(3) that they be studied in at least one representative species of each 
recognized supraspecific taxon; 

(4) that they not require extensive quantitative analysis. This restriction 
was imposed primarily because of the lack of series of a sufficiently diverse 
group of representative species large enough to support a reasonable degree 
of quantitative rigor; 

(5) that, except in cases of monobasic genera, they not be characters 
unique to single species. 

Even though most are easily observed, characters were not avoided be- 
cause they require special effort to assess. 

About 90% of all known species representing all phanaeine genera were 
examined during the course of this study, including most of the rarer ones. 
However, not all species were examined with the same thoroughness; the 
appendix indicates the extent to which each was examined. Since a revision 
of species is not the present problem, no special effort was made to obtain 
large series (more than 20 individuals). Such large series were studied in 
some instances only to aid in determining what characters tended to vary 
greatly intraspecifically. Except in cases of extremely rare species (e.g., 
Homalotarsus impressus), the minimum number of specimens examined of 
each was an individual of each sex; usually more specimens were available. 

In general, the sequence of characters treated here is the same as in the 
foregoing section. The name of some characters is preceded by a number 
(1-22) to permit easy cross reference in following sections. In a few in- 
stances (e.g., character 6) certain highly correlated features are treated 
together even though they may be in quite distinct anatomical locations. 
Numbered characters are "generic characters" in that they generally have 
wide taxonomic application at the genus level; hence, they form the nucleus 



778 



The University Science Bulletin 




Figs. 79-82. 79, Phanaeus vindex, female, frontal view of head; 80, Oxysternon curvispinum, 
male, dorsal view of head; 81, Phanaeus haroldi, female, dorsal view of head; 82, P. melampus, 
female, dorsal view of head. 



of the generic descriptions given below. It should be said, however, that 
description of genera is limited neither to states of these numbered characters 
nor to the variation discussed in this section. 

The states of many characters are designated by a, b, etc. In these in- 
stances the discussion of each state of the character is introduced by a few 
"key" words followed, if necessary, by a complete description of the state. 
For brevity, these key words, which are not necessarily self-explanatory, 
will be used in subsequent discussions (keys, descriptions, etc.). If a given 
character state is reasonably regularly distributed at or above the subgenus 
level, its definition is followed by the names of those taxa possessing it. For 
this purpose, it is necessary to use names not formally introduced and de- 
fined until later; to gain some taxonomic perspective, one can refer to the 
classification outline presented at the beginning of Part V. Names of taxa 
used in this section may be followed by (ex) or (inf). When (ex), ex- 
ception (s), is noted, one or more species (but usually only a small per- 
centage of all species) of the cited taxon possess another defined state; unless 
they are pertinent to the discussion of the character, these exceptional species 



Evolution of the Phanaeine Dung Beetles 



779 





94 





Figs. 83-97. Anterior clypeal margins, dorsal views. 83, Copiophanaetis lancijer; 84, C. 
pertyi; 85, C. saphirinus; 86, C. dardanus; 87, C. pinto; 88, C. jasius; 89, Dendropacmon 
renatii; 90, D. viridis; 91, D. ganglbaueri; 92, D. amyntas; 93, D. bahianus? 94, D. 
denticollis; 95, Tetramereia convexa; 96, Megatharsis buchjeyi; 97, Homalotarsus impresstts. 



are not mentioned. When (inf), inferred, is noted, the character state in 
question is one requiring dissection to observe, for which expendable 
material of the cited taxon was not available. 

Discussion of some characters is omitted from this part. Those concern- 
ing color and size will be mentioned in the systematic section. For com- 
parison of the genital capsules of males, which were not examined for this 
study, the reader is referred to Olsoufieft's monograph (1924), particularly 
to Plates III-VII which illustrate the genital capsules of a large number of 
phanaeines. 

The Head Capsule 

Various aspects of the variation in structure of the head capsule provide 
important taxonomic characters at supraspecific levels. 

clypeal sculpturing: The entire dorsal surface of the clypeus of most 
phanaeines is completely transversely ridged (Figs. 79, 82). In some species 



780 



The University Science Bulletin 






Figs. 98-110. Anterior clypeal margins, dorsal views. 98, Diabroctis mimas; 99, D. 
mirabilis; 100, Oxystemon conspicillatum; 101, 0. palaemon; 102, Sulcophanaeus imperator; 
103, S. velutinus; 104, S. carnifex; 105, S. menelas; 106, Phanaeus splendidiilus; 107, P. 
endymion; 108, P. palaeno; 109, P. demon; 110, P. amethystinns. 



of Oxystemon (including all Mioxysternon, Fig. 80) and in isolated species 
in other genera it is punctate posteriorly. 

1. shape of the anterior margin of the clypeus. 

a. emarginate medially — interrupted medially by three (rarely one, 
Fig. 90) usually acute emarginations, middle one being deepest; emargina- 
tions setting off two large, acute median teeth (Figs. 83-89, 91-92). Copro- 
phanaens, Dendropaemon. In some species of Dendropaemon the clypeal 
teeth and emarginations are rounded (Figs. 93-94). The emarginate clypeus 
is usually coincident with the U-shaped clypeal process (see character 2a, 
below). 



Evolution of the Phanaeine Dung Beetles 



781 



111 Xf 



113 



B 





114 




Figs. 111-114. Clypeal processes, A — frontal view, B — lateral view, C — cross section. Ill, 
Phanaeus mexicanus; 112, Oxysternon conspicillatum; 113, Sulcophanaeiis jaitntts; 114, 
Coprophanaens spitzi (A — ventral view; B — frontal view). 



b. not emarginate medially — not interrupted medially by acute emar- 
ginations; medially rounded (Fig. 109) or distinctly bidentate (Figs. 95-108, 
110). All other genera not mentioned in a. This state is coincident with 
clypeal processes of several types (character 2, b-d) but never with a U-shaped 
process. 

Character 1 is the first of several (see section on phylogeny) indicating a 
presumed phyletic separation of Coprophanaeus and Phanaeus (and allied 
genera). At least in Coprophanaeus an emarginate clypeus, strengthened by 
a unique U-shaped clypeal process, as well as strongly toothed front tibiae 
(see character 18) are probably related to necrophagy (see part IV). The 
unique clypeal (and tibial) dentition of Coprophanaeus is approached by 
Phanaeus endymion and closely related species (Figs. 108, 256). 

2. clypeal process: 

a. U-shaped — curved ridge extending onto ventral surfaces of median 
clypeal teeth (Fig. 114, u). Coprophanaeus, Dendropaemon. Although usu- 
ally distinct, the process of Dendropaemon is only weakly produced. 

b. transverse— straight, transverse ridge usually wider than high (Figs. 
9-10, clypr; 111) but sometimes tooth-like (Fig. 113) or acute. Diabroctis, 
Sulcophanaeiis, Phanaeus (ex), Mioxystemon (very weakly produced). 

c. spiniform — laterally compressed, acute spine (Fig. 112). Some Oxy- 
sternon, sen. str. 

d. reduced — no distinct ridge, etc., present; process at most represented 
by a weak gibbossity, raised area or poorly defined ridge. Tetramereia, 
Homahtarsits, Megatharsis. In all three genera the ventral surface of the 



782 The University Science Bulletin 

clypeus is strongly swollen in the area occupied by the clypeal process in other 
groups; this swelling is similar to that bearing the U-shaped process. 

It should be noted that since the head functions as a digging instrument, 
such features as the anterior margin of the clypeus and clypeal processes are 
subject to wear which can alter their shape. 

lateral clypeal carinae : These ridges are usually present although their 
extent and distinctness is subject to variation which cannot meaningfully be 
partitioned into states. These carinae are characteristically absent in only a 
few supraspecific taxa {Mioxystemon, Tetramereia and Homalotarsus) as 
well as isolated species in nearly all other groups. 

The transverse clypeal carina is unique among phanaeines to Diabroctis 
(Figs. 194, 197, tcrcly); a similar carina is characteristic of Onthophagus 
(Onthophagini). 

3. cephalic process of the male: The shape of the male cephalic process 
varies greatly among species, in addition to intraspecific allometric variation. 
The following three types are the most common; they do not represent 
mutually exclusive groups since intermediate and unique types are not con- 
sidered. Except when otherwise indicated, the following apply only to large, 
well developed ("major") individuals; the type of process borne by smaller 
("minor") individuals, while reduced in size, is usually recognizable. 

a. lamellate — antero-posteriorly flattened process with or without two 
or three distal projections of equal or unequal size and shape; usually pre- 
senting flattened anterior surface (Figs. 186, 188, 192-193, 305). This type is 
characteristic of Coprophanaeus, sen. str., some Dendropaemon, and Dia- 
broctis mimas and cadmas. 

b. carinate — a transverse carina, wider than high. Most Dendropae- 
mon, Tetramereia, Homalotarsus (inf), Megatharsis (inf). 

c. corniform — a long, usually gradually tapering "horn" erect or curv- 
ing posteriorly over pronotum (Figs. 1, 181, 183, 189, 198, 203-204, 209-211, 
216, 218, 221, 226). This is by far the most common type and only type 
found in large males of the following taxa: Sidcophanaeus, Oxysternon, sen. 
str., Phanaeus, Megaphanaeus, Metallophanaeus. In small males of Oxy- 
sternon, sen. str., all Mioxystemon and two Phanaeus (palaeno and \irbyi) 
the process is reduced to a pair of closely set, median tubercles (Fig. 80). 

4. cephalic process of the female: In general, the female cephalic process 
is subject to much less interspecific and intraspecific variation than is that of 
the male. 

a. corniform — a long, tapering "horn" (Figs. 182, 199) or large, horn- 
like process (Fig. 201). This is an uncommon condition among phanaeines 
and other dung beetles; in general, these females are of species the males of 



Evolution of the Phanaeine Dung Beetles 



783 




Figs. 115-124. Head and prothoracic structure; A — lateral view of head region around eye, 
B — same, dorsal view, C — left side of anterior portion of pronotal margin, dorsal view. 115, 
Diabroctis mimas; 116, Coprophanaens jasius; 117, Sulcophanaeus jaunus; 118, Oxysternon 
conspicillatum; 119, Phanaeus haroldi; 120, Coprophanaetis saphirinus; 121, Sulcophanaeus 
chryseicoUis; 122, S. imperator; 123, Dendropaemon renatii; 124, D. jractipes. 



784 The University Science Bulletin 

which also have corniform processes; since the pronotal structure of the fe- 
male is usually also masculine, the sexes tend to closely resemble each other 
in general appearance. Corniform or corniform-like processes are character- 
istic of Megaphanaeus and the faunas group of Sulcophanaeus. 

b. carinate — a straight or anteriorly bowed, transverse carina, usually 
trituberculate (Figs. 81, 185, 197, 207, 213). This is the usual type of process 
in phanaeines and other dung beetles; it is characteristic of all phanaeine 
taxa except those mentioned above. In many species of Phanaeus it is a 
short thickened ridge which is more-or-less isolated between the eyes (Fig. 
82) and, in a few species (e.g., vindex, Figs. 3, 79), distinctly raised. 

paraocular areas of the parietals : These vary in many respects, among 
them sculpturing (smooth to strongly rugose) and relief. In Dendropaemon 
they are always strongly raised medially (above) as a large bump or thick 
transverse ridge (Fig. 134) ; they are flat in Tetramereia, Homalotarsus and 
Megatharsis. In other taxa they vary among species (convex, flat or con- 
cave). In most phanaeines the paraocular areas are carinate lateral to the 
upper portion of the eye; this carina is characteristically lacking in few multi- 
specific taxa (e.g., Coprophanaeus, sen. sir.). These areas also vary in width, 
the basis of the following character: 

5. width of the paraocular areas: 

a. narrow — width of area, along its posterior margin, less than three 
times (usually about two times) width of adjacent dorsal portion of eye 
(Figs. 115B-123B). All taxa except those mentioned in b. 

b. wide — width at least three times (usually more) width of eye (Figs. 
124B, 134-135). Dendropaemon, sen. str., (ex), Tetramereia, Homalotarsus, 
Megatharsis. 

6. postocular lobes of the parietals : The shape of the posterior portion 
of the head capsule varies in a number of rather subtle ways, most of which 
indicate no regular taxonomic pattern. The most easily described of these 
variables is the angle at which the postocular lobes of the parietals are de- 
pressed posterior to the dorsal portions of the eyes (as seen dorsally). Al- 
though intermediates occur, the state of the members of most taxa is clearly 
one of two extremes, each correlated with a distinct shape of the anterior 
portion of the circumnotal ridge: 

a. depressed transversely — lobes depressed straight behind eyes in plane 
perpendicular to long axis of body (Figs. 115B-117B, 120B-124B) ; anterior 
portion of circumnotal ridge not broken by emarginations receiving the 
postocular lobes (same Figs., C). All genera except Phanaeus and Oxy- 
sternon. In the dardanus group of Coprophanaeus, sen. str., the circumnotal 
ridge is effaced behind the eyes but not distinctly emarginate. In all Copro- 



Evolution of the Phanaeine Dung Beetles 



785 








¥ A 



~l 130 





f ^fu 



131 




Figs. 125-131. Head structure; A — inner view of cephalic brachium and adjacent structures 
(stippling indicates surrounding intracephalic space), B — cross section through brachium, 
C — outer view of head wall in region of mesal branch of fronto-clypeal sulcus (indicated by an 
arrow). 125, Sulcophanaeiis f annus; 126, Coprophanaetis jasius; 127, Phanaeus demon (female); 
128, Oxysternon conspicillatum; 129, O. palaemon; 130, Phanaeus palaeno; 131, P. chalcomelas. 

phanaeus and Dendropaemon the postocular lobes are conspicuously flat- 
tened behind (compare polb in Figs. 115A and 116A). 

b. depressed obliquely — lobes depressed behind eyes in plane oblique to 
long axis of body; depressions setting ofT prominent, posteriorly directed 
portions of the lobes (Fig. 4) received by distinct emarginations (Fig. 35, 
emarg) interrupting anterior position of circumnotal ridge. Phanaeus (ex). 
Oxysternon, sen. str., characteristically shows a state intermediate to a. and 
b. above (Fig. 118); the circumnotal margin, while usually weakened by 
shallow emarginations, is not always distinctly broken; in Mioxysternon the 
entire anterior portion of the circumnotal ridge is effaced. 

canthal area: The shape and extent of this area varies more or less 
continuously with most taxa possessing one of two extremes: 



786 The University Science Bulletin 

a. distinct — posterior margin of paraocular area (as seen dorsally) 
approaching middle of eye, leaving, as seen laterally, a flattened or concave 
canthal area (Figs. 115A, ca, 116A-118A, 121A-124A) between upper and 
lower portions of eye. Most members of all taxa except Phanaeus and 
Mioxystemon. 

b. indistinct — posterior margin of paraocular area approaching poste- 
rior angle of eye, not leaving a distinct canthal area (Figs. 1, 4). Mioxy- 
stemon and Phanaeus. 

Figure 120A shows an intermediate state. 

7. SHAPE OF CEPHALIC BRACHIUM: 

a. bipodal — presenting two distinct "arms" (Figs. 5, cbr; 128, 131); 
mesal branch of fronto-clypeal sulcus Y- or T-shaped (Fig. 1, mbfcs). 
Phanaeus, Oxystemon, sen. str., Mioxystemon (inf). In a few species of both 
genera the bipodal nature is obscured by a presumed reduction in size of one 
or both "arms" (Figs. 127A, 129A, BOA) and coincident shortening or loss 
of the anterior extension of the sulcus (same Figs., C) ; in these cases only 
cross sections of the respective brachia (same Figs., B) show that they are 
bipodal. 

b. monopodal — presenting a single, usually straight and vertical "arm" 
(Fig. 126A); mesal branch of fronto-clypeal sulcus not forked, usually 
straight (Fig. 126C). All genera except Phanaeus and Oxystemon; inferred 
for Homalotarsus. The brachia of Sulcophanaeus faunus (Fig. 125A) and 
S. carnijex are apparently unique monopodal types which appear super- 
ficially as bipodal (a few Sulcophanaeus do have bipodal brachia). As indi- 
cated by the course of the mesal branch of the fronto-clypeal sulcus (Fig. 
125C) and cross section of the brachium (Fig. 125B), the brachium of these 
species is a single, anteriorly bent vertical plate. 

Since the shape (forked or not) of the mesal branch of the fronto-clypeal 
sulcus can be considered a projection of a cross section of the brachium, type 
of brachium can usually be determined accurately without internal exami- 
nation. The sulcus is regularly obscured only in the auricollis group of 
Sulcophanaeus. 

8. size of eyes: 

a. small — dorsal interocular width more than four times width of dor- 
sal portion of an eye (Figs. 4, 134-135) ; width of lower portion of an eye less 
than twice (usually about equal) width of adjacent oculo-gular space (Fig. 
9). With few rare exceptions, all taxa except Megaphanaeus, Coprophanaeus, 
sen. str., and Coprophanaeoides. 

b. large — dorsal interocular width less than four times width of dorsal 
portion of an eye (Fig. 132) ; width of lower portion of an eye at least twice 



Evolution of the Phanaeine Dung Beetles 



787 




*rW- — ^fy- 




133 





Figs. 132-135. 132, Dendropaemon renatii, dorsal view of anterior portion of head; 133, 
same, ventral view of posterior portion of head; 134, D. viridis, dorsal view of anterior portion 
of head; 135, Megatharsis buckjeyi, dorsal view of anterior portion of head. 



(usually more) width of adjacent oculo-gular space (Fig. 133). Mega- 
phanaens, Coprophanaens, sen. str. and Coprophanaeoides. 

9. occipital ridge: 

a. complete — usually setose ridge extending dorsally over posterior po- 
sition of head (Fig. 1, 4, 8, ocr) ; small medial section occasionally obscured 
by punctures. Oxysternon, Phanaeits. 

b. incomplete — ridge absent or, at most, represented by usually weak 
lateral segments, never distinct dorsally. All genera except Oxysternon and 
Phanaetts. 

In all phanaeines except Megaphanaeus the occipital areas are flattened or 
weakly concave; in this subgenus, however, the areas are abruptly raised 
postero-laterally as prominent bumps, anterior to which they are strongly 
concave. 

Appendages of the Head 

antennae: The antennae are subject to some subtle variation which is 
exceedingly difficult to render as distinct character states without the use of 
statistical procedures (e.g., relative lengths of antennomeres, etc.) ; very little 
antennal variation is striking enough to assess otherwise. Noteworthy among 
this latter variation is the tendency in Dendropaemon for elongation of first 



788 



The University Science Bulletin 







137 



138 



139 



Figs. 136-139. 136, Dendropaemon jractipes, flagellum of right antenna, dorsal view; 137, 
D. denticollis, A — apical view of left antennal club, B — apical five segments of left antenna, 
ventral view; 138, Homalotarsus impressus, apical five segments of right antenna, dorsal view; 
139, Megatliarsis buchleyt, flagellum of right antenna, dorsal view. 



lamella (Fig. 136) and dorse-ventral flattening of the entire club (Fig. 137) 
and the near hemispherical club of Megatliarsis (Fig. 139) and Homalotarsus 
(Fig. 138). 

labrum: Variation in the labrum offers a number of important taxo- 
nomic characters, both unique to few taxa and with wide application. 
Among the unique features are two characteristic only of Dendropaemon: 
(1) apical margin deeply emarginate (Fig. 145) (that of other phanaeines 
may be weakly emarginate), and (2) antero-median process of suspensorium 
forked anteriorly, within fork insert setae of median brush (Fig. 145). Labral 
characters of potential significance which were not studied include various 
aspects of chaetotaxy (e.g., number of setae in lateral files) ; study of these 
was not approached since their efficient interpretation would have required 
statistical methods (see introduction to this part). 

One of the principal aspects of labral variation is the definition of two 
principal labral types based on the correlation of three characters: lateral 
files, type of median brush and type of labral suspensorium. 

10. LABRAL TYPE: 

a. Coprophanaeus-type — lateral files of labrum parallel or nearly so 
(divergent in Coprophanaeits), distance between anterior ends less than 1.25 
(usually about 1.0) times distance between posterior ends (Figs. 140, 142, 
144) ; central setae of median brush much longer than setae posterior to 
them (Fig. 148) or median brush comprising only few, very long setae (Fig. 
144) ; length of opening in suspensorium 0.5 times or less length of dorsal 
surface of labrum, opening usually very transverse, formal arms narrow and 
often converging posteriorly (Figs. 141, 143, 145-146). Coprophanaeits, 
Homalotarsus (inf), Tetramereia, Megatliarsis, Dendropaemon. These 



Evolution of the Phanaeine Dung Beetles 



789 




140 




142 




% • «y v$ 



■ 



144 




141 








143 




145 



146 





Figs. 140-148. Labra. 140, Megatharsis bachjeyi, ventral view; 141, same, dorsal view; 
142, Tetramereia convexa, ventral view; 143, same, dorsal view; 144, Dendropaemon viridis, 
ventral view; 145, same, dorsal view; 146, Coprophanaeus jasius, dorsal view; 147, same, ventral 
view; 148, same, lateral view of median brush of labrum. 

genera, in the foregoing order, show progressively a reduction in the number 
of setae comprising the median brush, from a large number in Copro- 
phanaeus to a few in Dendropaemon; the labral suspensorium of the last four 
genera is weakly sclerotized compared to that of other phanaeines. 

b. Phanaeus-type. — lateral files divergent, distance between anterior 
ends more than 1.25 (usually about 1.5) times distance between posterior 
ends (Figs. 19A, 147) ; central setae of median brush not much longer than 
setae posterior to them (Fig. 18), brush never reduced to a few setae, long or 



790 



The University Science Bulletin 




Figs. 149-158. 149, Sttlcophanaeus batesi, dorsal view of labium (floor of buccal cavity not 
shown); 150, same, lateral view (slightly smaller scale than Fig. 149; only apical portion 
shown in detail); 151, Dendropaemon viridis, lateral view of labium (only apical portion 
shown in detail); 152, same, ventral view of mentum and palpi; 153, Tetratnereia convexa, 
ventral view of mentum and palpi; 154, Megatharsis buckjeyi, ventral view of mentum and 
palpi; 155, same, dorsal view of right palpus; 156, Homalotarsits impressus, ventral view of 
mentum and palpi; 157, Coprophanaens jasitis, ventral view of mentum and palpi; 158, 
Sulcophanacus jaunus, ventral view of mentum. 



Evolution of the Phanaeine Dung Beetles 791 

not; length of opening in suspensorium more than 0.5 (usually nearly 1.0) 
times length of dorsal surface of labrum, opening rounded, not distinctly 
transverse, tormal arms thick (Fig. 19A). Diabroctis, Sulcophanaetis, 
Phanaeus, Oxysternon. 

mandibles and maxillae: These mouthparts are remarkably similar in 
all taxa studied; they yielded no taxonomic characters useful at the supra- 
specific level. 

labium: Unique variation of the labium is restricted to Megatharsis 
buc^leyi, which, unlike any other phanaeine, has one-segmented labial palpi 
(Fig. 154) ; the dorsal surface of each segment bears a membranous slit (Fig. 
155, arrow) possibly indicating an incomplete line of fusion. Other labial 
characters, with wider taxonomic application, include the following three: 

11. SHAPE OF MENTUM: 

a. weakly emarginate — anterior emargination shallow or scarcely dis- 
tinct, its depth less than l / 4 (with few exceptions l / 8 or less) greatest length 
of mentum; greatest width of mentum usually less than 1.5 (approaching 1.0 
in Dendropaemon) times greatest length (Figs. 152-154, 156-157). Copro- 
phanaeus, Dendropaemon, Tetramereia, Homalotarsus, Megatharsis. 

b. strongly emarginate — anterior emargination deeper, depth at least 
! /4 (usually Yi or more) length of mentum; width of mentum almost always 
1.5 or more times length (Figs. 25, 158). Diabroctis, Sidcophanaeus, 
Phanaeus, Oxysternon. 

12. MEDIAN LOBE OF HYPOPHARYNX : 

a. present — large, distinct lobe at summit of premental groove (Fig. 
26, mlhpx). All genera except those mentioned below. 

b. absent — no distinct lobe present. Dendropaemon, Tetramereia, 
Homalotarsus (inf), Megatharsis. 

13. PREMENTAL SCLERITES : 

a. completely sclerotized — entire sclerite darkly sclerotized (Fig. 30, 
pmts). Phanaeus, Oxysternon. 

b. marginally sclerotized — only a C-shaped portion of posterior mar- 
gin darkly sclerotized, otherwise weakly sclerotized (Figs. 150-151, pmts). 
All genera except Phanaeus and Oxysternon. 

There exists another labial character — of great interest but little taxo- 
nomic value at the supraspecific level — concerning the position of the glossae, 
which varies between the following extremes: 

a. appressed. — glossae closely appressed along mesal margins for at 
least basal half of their lengths, completely closing over median groove of 



792 



The University Science Bulletin 




Figs. 159-162. 159, Phanaeus pyrois, male, left side of pronotum; 160, Oxystanon conspicilla- 
tum, male, same; 161, Phanaeus haroldi, female, same; 162, Dendropaemon viridis, caudal 
view of pygidium. 



prementum and enclosing between their bases median lobe of hypopharynx 
(if present) (Fig. 26). 

b. separated — glossae widely separated basally, not touching mesally 
at any point; median groove of prementum exposed, median lobe of hypo- 
pharynx (if present) free (Fig. 149). 

Species showing the above extremes and intermediate positions are in- 
cluded in each of the multispecific genera except Oxysternon, all examined 
species of which have appressed glossae. The following, for each genus, lists 
three common species showing, respectively, appressed, intermediate and 
separated glossae : 



Evolution of the Phanaeine Dung Beetles 



793 




Figs. 163-166. 163, Phanaens damocles, female, antero-lateral portion of pronotum; 164, 
P. triangularis, female, same; 165, P. vindex, female, postero-median portion of pronotum; 
166, Coprophanaeus jasius, female, left antero-lateral portion of pronotum. 



Phanaens: vindex, palaeno, qitadridens 
Sidcophanaens: f annus, imperator, batesi 
Coprophanaeus: ensifer, jasius, milon 
Dendropaemon: viridipennis, denticollis, renatii 
The glossae of Megatharsis are appressed; those of Tetramereia and 
Homalotarsus (inf ) are of intermediate position. 

Thoracic Segments 
The shape of the pronotum of most phanaeines differs markedly between 
the sexes and among species. Like those of the head, pronotal outgrowths 
usually vary allometrically within single local populations. The form of the 



794 



The University Science Bulletin 




Figs. 167-180. 167, Dendropaemon amyntas, male, lateral view of pronotum; 168, D. 
renatii, female, same; 169, same, dorsal view of pronotum; 170, D. riridis, male, lateral view 
of pronotum; 171, same, dorsal view of pronotum; 172, Tetramereia convexa, lateral view of 
pronotum; 173, Dendropaemon jractipes, same; 174, D. denticoUis, same; 175, D. haroldi, same; 
176, D. nitidicollis , same; 177, same, dorsal view of pronotum; 178, Megatharsis buc\leyi, lateral 
view of pronotum; 179, D. haroldi, ventral view of left side of prothorax and base of leg; 180, 
D. viridipennis ,, same. 



Evolution of the Phanaeine Dung Beetles 



795 




Figs. 181-189. 181, Coprophanaetts lancifer, male, lateral view of head and pronotum; 182, 
same, female; 183, C. saphirinus, male, same; 184, same, female, lateral view of pronotum; 185, 
C. pertyi, female, dorsal view of head and anterior portion of prothorax; 186, C. pluto, male, 
lateral view of head and pronotum; 187, same, dorsal view of median pronotal prominence; 
188, same, anterior view of cephalic process; 189, C. pertyi, male, lateral view of head and 
pronotum. 



male pronotum varies greatest among species and is often distinctive of a 
species. Female pronotal form varies less among closely related species; 
indeed, the females of some groups of species (e.g., Phanaeus daphnis and 
related species) are impossible to distinguish using secondary sexual charac- 
ters of the head and pronotum, even though the males may be quite distinct. 
It is difficult to arrange pronotal variation in shape into reasonably distinct 



796 



The University Science Bulletin 




Figs. 190-197. 190, Coprophanaeus jasius. female, dorsal view of anterior portion of prono- 
tum; 191, same, lateral view of pronotum; 192, same, male, lateral view of head and pronotum; 
193, same, male, anterior view of cephalic process; 194, Diabroctis tnimas, male, dorsal view 
of head and pronotum; 195, same, lateral view; 196, same, female, lateral view of pronotum; 
197, same, female, dorsal view of head. 



Evolution of the Phanaeine Dung Beetles 



797 




Figs. 198-203. 198, Sulcophanaeus faunus, male, lateral view of head and pronotum; 199, 
same, female: 200, same, female, dorsal view (incomplete) of pronotum; 201, S. carnifex, 
female, lateral view of head and pronotum: 202, same, female, anterior view of cephalic 
process; 203, same, male, lateral view of head and pronotum. 



character states. Obvious types do exist, although their taxonomic reliability 
is reduced by a fairly high incidence of presumed convergence, particularly 
among males. In general, however, closely related species have similar 
pronotal shape. 

Characters 14 and 15 treat pronotal types based on the shape and position 
of various projections, concavities, etc. Character 14 is based on well de- 



798 



The University Science Bulletin 




Figs. 204-209. 204, Sidcophanaeits imperator, male, lateral view of head and pronotum; 
205, same, female; 206, S. velutinus, male, lateral view of pronotum; 207, S. chrysekollis, 
female, dorsal view of head and anterior portion of pronotum; 208, same, female, lateral view 
of pronotum; 209, same, male, lateral view of head and pronotum. 



veloped ("major") individuals; less well developed males, while most often 
presenting reduced versions of the same features, may appear quite distinct. 
Like cephalic structure, variation in pronotal shape among conspecific fe- 
males is much less pronounced than among males; thus, the types defined 
in character 15 should apply equally well to large and small individuals. The 
following do not include unique types; furthermore, since the variation 
within types is sometimes quite wide, the defined types should not be 
considered strictly mutually exclusive. 

14. SHAPE OF PRONOTUM OF MALE: 

a. type A — postero-lateral angles produced laterally as rounded lobes, 
giving pronotum a decidedly triangular or trapezoidal shape as seen dorsally; 
usually flattened dorsally, though sometimes convex or concave and bearing 
one or more tubercles or stronger prominences on dorsal surface (Figs. 35-36, 
218-219,223-226,228). 

b. type B — bearing two latero-median processes separated by con- 
cavity (Figs. 210-211, 216-217). 

c. type C — strongly and widely raised antero-medially, bearing trans- 



Evolution of the Phanaeine Dung Beetles 



799 




Figs. 210-215. 210, Oxystemon conspicillatum, male, dorsal view of head and pronotum; 
211, same, lateral view; 212, same, female, lateral view of pronotum; 213, same, dorsal view 
of head and anterior portion of pronotum; 214, O. curvispinum , male, dorsal view of head and 
anterior portion of pronotum (separated); 215, same, lateral view of pronotum. 



verse angulate prominence or carina, anterior to which pronotal surface is 
vertical or concave (Figs. 167, 186, 189, 192, 194-195). 

d. type D — bearing four tubercles or prominences in transverse row- 
near posterior margin of pronotum; tubercles may differ little in size or 
inner two or outer two may be strong projections (Figs. 198, 203-204). 

e. type E — pronotum deeply and widely concave medially, posterior 
margin of concavity bearing massive, saddle-shaped prominence (Fig. 181). 

f. type F — flattened and sloping anteriorly with pair of closely set 
postero-median prominences which may be crest-like or elongate projections 
(Figs. 183, 206, 209). 



800 



The University Science Bulletin 




Figs. 216-222. 216, Phanaeus dejeani, male, lateral view ol head and pronotum; 217, same, 
dorsal view; 218, /'. endymion, male, lateral view of head and pronotum; 219, same, male, 
dorsal view of pronotum; 220, same, female, lateral view of head and pronotum; 221, P. 
haroldi, male, same; 222, same, female, dorsal view of head and anterior portion of pronotum. 



15. SHAPE OF PRONOTUM OF FEMALE: 

a. type A — convex except for antero-median concavity bordered an- 
teriorly by transverse carina or crest (Figs. 184-185, 190-191). 

b. type B — convex except for three closely set antero-median tubercles, 
median one of which may be dentate (Fig. 220). 

c. type C — convex except for low, antero-median prominence bordered 
anteriorly by unevenly bowed carina (Figs. 207-20S). 

d. type D — convex but bearing an antero-median transverse carina 
or wide ridge anterior to which pronotal surface sometimes nearly vertical 
(Figs. 37-38, 196, 205, 213, 227). 

e. type E — evenly convex (Figs. 178,310-315). 



Evolution of the Phanaeine Dung Beetles 



801 




Figs. 223-228. 223, Phanaeus lugens, male, lateral view of pronotum; 224, same, dorsal 
view; 225, P. demon, male, lateral view of pronotum: 226, P. mexicanus, male, lateral view of 
head and pronotum; 227, same, female, lateral view of pronotum: 228, P. daphnis, male, lateral 
view of pronotum. 



16. proxotal sculpturing: The texture of the pronotal surface varies con- 
siderably. Although intermediate conditions occur in some species most 
supraspecific taxa are characterized by one of the following states: 

a. smooth-punctate — completely smooth or with limited weakly 
roughened areas, but entire surface punctate (Figs. 159-161). 

b. roughened — surface entirely granulate (Fig. 164), granulo-rugose 
(Fig. 163) or rugose (Fig. 165); if distinct puncturing present, usually re- 
stricted to postero-median area. In most Coprophanaeus and Coprophanae- 
otdes the roughening, at least anteriorly, is in the form of transverse ridging 
(Fig. 166). 

The following are other taxonomically restricted variations of the 
prothorax : 

posterior proxotal fossae: These fossae are usually evident as puntiform 
pits or large rounded depressions on the postero-medial angle of the pro- 
notum (Fig. 194, pfos) ; they are characteristically absent only in Oxysternon, 
the postero-median angle of the pronotum of which is a strongly produced, 
acute process fitting between the bases of the elytra (Fig. 210). 



802 



The University Science Bulletin 




Figs. 229-245. 229, Dendropaemon denticollis, dorsal view of mesoscutellum; 230, same, 
lateral view of antero-median angle of metasternum; 231, same, ventral view of pterothorax; 
232, same, lateral view of metepisternum; 233, D. haroldi, same as 230; 234, same, same as 231; 
235, same, same as 232; 236, D. nitidicollis, same as 230; 237, same, same as 231; 238, same, 
same as 232; 239, Homalotarsits impressus, dorsal view of mesoscutellum; 240, same, ventral 
view of mesosternum and median area of metasternum; 241, Dendropaemon renatii, lateral 
view of metepisternum; 242, D. bahianus?, same; 243, D. hirticollis, same; 244, Oxysternon 
conspicillatum , lateral view of pterothorax (elytra in place, coxae removed); 245, same, ventral 
view of mesosternum and median area of metasternum. 



Evolution of the Phanaeine Dung Beetles 



803 




Figs. 246-259. Front tibiae. (All Figs, except 247B are of anterior surfaces.) 246, 
Coprophanaetts jasins; 247, Dendropaemon viridis (A — anterior surface, B — posterior surface) ; 
248, Megatharsis buchj.eyi; 249, Tetramereia convexa; 250, Diabroctis mimas (A — male, B — 
female); 251, Sulcophanaeus noctis, male; 252, S. columbi, female; 253, S. imperator, male; 
254, Homalotarsus impressus; 255, Phanaeus haroldi, female; 256, P. blanchardi, female; 257, 
P. \irbyi, male; 258, Oxysternon festivum, male; 259, Phanaeus mexicantts (A — female, B — 
male). 



804 



The University Science Bulletin 




Figs. 260-262. 260, Homalotarsus impressus, female, dorsal view of apex of femur, tibia 
and tarsus of right hind leg; 261, Phanaeus vindex, male, ventro-lateral view of left hind tibia 
and tarsus; 262, Sulcophanaeus j annus, male, same. 



lateral pronotal carinae : These carinae (Figs. 199, 201, lpc) are dis- 
tinct only in the Sulcophanaeus jaunus group. In other Sulcophanaeus, in 
Megaphanaeus and Diabroctis (particularly males) the pronotum is thick- 
ened along a non-carinate line in the same relative position as this carina. 
The lateral pronotal carinae of S. columbi are produced postero-laterally as 
small crests. These angles produce a pronotal shape similar to type A (char- 



Evolution of the Phanaeine Dung Beetles 



805 




263 




265 






Figs. 263-269. 263, Tetramereia convexa, ventral view of left hind tibia and tarsus; 264, 
Homalotarsus impressus, dorsal view of right hind tarsus; 265, same, ventral view of first 
segment of right mesotarsus; 266, Megatharsis buc\leyi, dorsal view of left hind tibia and 
tarsus; 267, same, dorsal view of middle left tibia and tarsus; 268, Dendropaemon viridis, dorsal 
view of right hind leg (except tarsus) ; 269, same, ventral view. 



acter 14), suggesting that the rounded postero-lateral lobes characteristic of 
type A are homologs of lateral pronotal carinae. 

transverse propleural carinae: These carinae are present only in Den- 
dropaemon, sen. str. When complete, each extends along the posterior mar- 
gin of the procoxal cavity onto the propleural surface lateral to the coxa (Fig. 
180, arrows); when incomplete, each is restricted along the posterior margin 
of the coxal cavity (Fig. 179, arrow). 

Variation in the pterothoracic segments of taxonomic significance is lim- 
ited to the spinate prolongation of the metasternum of Oxysternon (Figs. 



806 



The University Science Bulletin 







Figs. 270-279. 270, Dendropaemon denticollis, ventral view of right hind tibia and tarsus; 
271, D. haroldi, ventral view of right hind tarsus; 272, D. smaragdinum , same; 273, D. 
lobatum?, same; 274, D. rcnat'u, same; 275, D. amyntas, same (worn); 276, D. tiridipennis , 
same; 277, D. viridis, same; 278, D. ganglbaueri, same (arrow points to small second segment); 
279, same, ventral view of left middle tarsus. 



244-245) and to various aspects of body shape variation in Dendropaemon 
(Figs. 229-238, 241-243), which are discussed more fully in the systematic 
section. 

Legs 
17. front tarsi of female: 

a. present — short, five-segmented tarsi inserted on inner apical angle 
of tibia. Diabroctis, Sulcophanaeus (ex), Oxysternon, Phanaeus, Mega- 
phanaeus. 

b. absent — Coprophanaeus, sen. str., Metallophanaeus, Dendropae- 
mon, Tetramereia, Homalotarsus, Megatharsis. 



Evolution of the Phanaeine Dung Beetles 



807 




Figs. 280-283. Left elytron, dorsal views. 280, Oxysternon (M.) curvispinum; 281, O. 
conspicillatum (arrows point to second stria); 282, Sulcophanaeus noctis; 283, S. columbi 
(arrow points to third stria). 



18. FRONT tibiae: 

a. Coprophanaeus-type — very strongly and acutely quadridentate, 
teeth separated basally by very narrow spaces; apical three teeth always 
carinate (Fig. 246). Coprophanaeus. 

b. Dendropaemon-type — acutely quadridentate, teeth not separated 



808 



The University Science Bulletin 




Figs. 284-287. Lett elytron, dorsal views. 284, Phanaeus vindex; 285, P. mexicanus; 286, 
/'. damocles; 287, /'. endymion. 



by narrow spaces basally; apical three teeth always carinate; inner basal angle 
expanded, covering apex of femur (Fig. 247). Dendropaemon (except 
ganglbaueri). 

c. Phanaeus-type — tridentate or quadridentate, teeth acute or rounded, 
separated basally by wide, usually rounded spaces; apical two or three teeth 



Evolution of the Phanaeine Dung Beetles 



809 




Figs. 288-291. Left elytron, dorsal views. 288, Coprophanaeus lancijer; 289, C. jasius; 
290, Dendropaemon vindis; 291, same (higher magnification to show carinulate striae). 



carinate (Figs. 248-259) ; tibiae of male sometimes distinctly narrower than 
those of female (Figs. 250, 259). All genera except above two. 

19. MIDDLE AND HIND TARSI : 

a. two-segmented — some Dendropaemon, sen. sir., Paradendropae- 
mon (Figs. 272, 277-279). 



810 



The University Science Bulletin 










Figs. 292-299. (Stippling indicates areas for which detail not shown.) 292, Coprophanaetis 
jasius, right hind wing; 293, Sulcophanaeus chryseicollis , base of left hind wing; 294, 
Dendropaemon renatii, same; 295, D. piceum, same; 296, Megatharsis buckleyi, same; 297, 
Tetramereia convexa, same; 298, Dendropaemon haroldi, same; 299, D. amyntas, same. 



b. three-segmented — some Dendropaemon, sen. str., Coprophanae- 
oides (Figs. 270-271, 273-276). 

c. four-segmented — Tetramereia (Fig. 263). 

d. five-segmented — all other genera; relative shapes and sizes of seg- 
ments usually as in vindex (Fig. 64) ; shapes of tarsal segments of Mega- 
tharsis (Figs. 266-267) and Homalotarsus (Figs. 264-265) unique. 



Evolution of the Phanaeine Dung Beetles 



811 



300 





spr s 





Figs. 300-303. 300, Dendropaemon ganglbaueri, female, ventral view of abdomen; 301, 
Megatharsis buc\leyi, female, caudal view of abdomen (setae shown on left side only); 302, 
Sulcophanaeits carnifex, lateral view of right dorso-lateral extremity of abdominal sternum 7 
and adjacent areas; 303, Coprophanaetts dardanus, same. 



transverse tibial carina: In addition to a few isolated species (e.g., 
Sulcophanaeits imperator), this carina (Fig. 262, cr) is restricted to the Sulco- 
phanaens faunus group, Diabroctis, and the Oxysternon festivum group. 

Other variation of the legs includes the following, relatively isolated spe- 
cializations: The setigerous granulation of the dorsal surface of the hind legs 
of Homalotarsus (Fig. 260) and the characteristic expansion of the lower 
apical angle of the middle and hind femora of Dendropaemon (Figs. 268- 
269). 

Wings 
20. sculpturing of elytral interstriae : 

a. smooth — no distinctly punctured or rugose areas; if punctures pres- 
ent, restricted to minute punctures clearly visible only with high (20X) 
magnification (Figs. 281-282, 285, 2S7, 289). 

b. punctate — distinctly punctured, punctures visible with low (10X) 
or no magnification (Figs. 280, 286, 290). 

c. rugose — granulate, rugose or very heavily punctured with co- 
alescing of punctures (Figs. 284, 288). 



812 The University Science Bulletin 

21. relief of elytral interstriae : 

a. flat— (Figs. 282, 284, 286). 

b. convex— usually evenly (Figs. 280-281, 285, 287, 290-291) but some- 
times raised only medio-longitudinally such that striae appear to lie in broad, 
shallow troughs (Fig. 283). 

22. striae: 

a. carinulate — bordered laterally by fine carinae which may be straight 
(Figs. 290-291) or undulate (Fig. 288). Megaphanaeus, Dendropaemon (ex- 
cept ganglbaueri) and isolated instances in other genera. 

b. simple — distinct or very weak, sometimes punctate lines with un- 
defined lateral margins (Figs. 280-287, 289). All taxa except above. 

Of the characters of the hind wing only one is reasonably consistent at 
the supraspecific level: a distinct notching of the wing between the vannal 
and jugal lobes (Figs. 292, 294-295, 298) characteristic of Coprophanaeus, 
sen. str., Coprophanaeoides and some Dendropaemon, sen. str.; all others 
(except Phanaeus bispinus) have no emargination (Figs. 71, 293, 296, 299). 

Abdomen 
Only the following features comprise marked departure from the other- 
wise striking consistency in abdominal structure among the phanaeines 
studied : 

position of last (7th) abdominal spiracle: The opening of this spiracle 
assumes three positions. 

a. marginal — spiracular opening located in membrane connecting 
tergum 7 and sternum 7 (Fig. 72, spr?); usually visible only dorsally; this is 
the most common position. 

b. intermediate — located in a notch in dorsal margin of sternum 7; 
usually visible laterally (Fig. 303). 

c. intrasternal — located within dorso-lateral portion of sternum 7 (Fig. 
302, spr), faint suture (s) extending from spiracle to dorsal margin of 
sternum 7; visible only dorsally. 

Supposedly the evolution of position of this spiracle has included its ven- 
tral migration and/or dorsal extension of sternum 7, ultimately becoming 
isolated by closure of the sternum along a distinct suture. The intrasternal 
position is the least frequent and characteristic only of Megaphanaeus and 
Mioxysternon as well as isolated species-pairs in Sidcophanaeus. States a. and 
b. occur in a random taxonomic pattern. 

The only other noteworthy abdominal variations are the lateral depres- 
sions of the base of the pygidium of Dendropaemon (Fig. 162, Id), the 
sternal fringes of Megatharsis (Fig. 301) and the abnormally elongate 
sternum 7 of Paradendropaemon (Fig. 300, s?). 



Evolution of the Phanaeine Dung Beetles 813 

PART IV: ADAPTIVE SIGNIFICANCE OF ASPECTS OF 
PHANAEINE SKELETAL MORPHOLOGY 

What is known about the habits of phanaeines, as well as other Scara- 
baeinae, has been well summarized by Halffter and Matthews (1966). One 
important group of adaptive modifications of dung beetles, those related to 
coprophagy, has been discussed by them (pp. 241-248). As they and others 
have pointed out, adult scarabaeines are basically fossorial animals. Al- 
though they seek food on the surface, a large part of the feeding, mating 
and reproductive activities is carried on in underground tunnels and cham- 
bers excavated by the beetles. Efficient fossorial habit requires a highly 
specialized morphology. The purpose of this discussion is to bring out 
certain aspects of adult scarabaeine morphology, as illustrated specifically 
by the phanaeines, presumably related to digging habits. Certain characters 
here deemed fossorial are no doubt also important in other behavioral con- 
texts; indeed, a portion of this discussion is the examination of just such a 
case. Nevertheless, it is assumed that the initial selective factors involved in 
their evolution from antecedent types favored morphological complements 
of fossorial behavior. 

The fossorial features of phanaeines, and dung beetles in general, fall 
into two general categories: (1) structural modifications of direct use as 
digging tools, and (2) strengthening devices for increased body resistance 
to the generally compressive forces exerted on the body during digging and 
manipulation of soil (and food). Most of the features to be mentioned are 
shared with other Scarabaeinae and some other laparostict scarabaeids. 

Structural Tools 
The primary earth-moving structures are the head and front legs. The 
following modifications of the head capsule and legs of phanaeines pre- 
sumably contribute to the efficiency of these organs as digging tools: 

a. Differential expansion and ventral folding of the head sclerites — This 
trend has two very important adaptive consequences, (1) increase in the 
digging surface area of the head, principally through dorsal expansion of 
the clypeus, and (2) elimination of dorsal and lateral exposure of the head 
appendages, which are thereby protected from potential injury. 

b. Clypeal dentition and transverse ridging — While the dorsum of the 
head (clypeus) is primarily employed in pushing and packing soil (and 
food), its anterior margin functions primarily as a scraper. Clypeal teeth 
and other irregularities produce sharper, thus more effective cutting instru- 
ments. In addition to teeth, the transverse ridging of the dorsal surface 
contributes to its overall abrasiveness. 

c. Antero-lateral flattening and lateral toothing of the front tibiae — The 
front tibiae are principally scraping structures whose effectiveness is in- 
creased by dentate cutting edges. 



814 The University Science Bulletin 

d. Elongation, double articulation and transverse orientation of the front 
coxae — Elongation has effectively increased the surface area of the coxal 
margin available for muscle insertion. If increased musculation is to func- 
tion efficiently to rotate the coxa, the axis of rotation of a subcylindrical coxa 
must be fixed by two articulations, one at each end. Transverse orientation 
of the coxa preserves an antero-posterior movement of the distal portion of 
the leg effective both for walking and digging movements (dirt is pushed 
under body). 

e. Strong development and musculation of middle and hind legs — These 
legs, while not directly involved in digging, propel the body forward and, 
to a large extent, determine the effectiveness of digging movements of the 
head; they are strongly musculated. The tarsi are articulated so that, during 
propulsion of the body by middle and hind legs, they project laterally to 
allow firm implantation of the expanded distal ends of the tibiae. 

Strengthening Devices 
To effectively aid digging movements, the body wall must provide rigid 
mechanical support. The following features (in addition to thickly sclero- 
tized integument) presumably contribute to the overall body resistance 
necessary to withstand the stresses imposed by digging: 

a. Cephalic brachium — While the brachium no doubt imparts necessary 
rigidity to the dorsal mandibular articulation, it also increases the resistance 
of the head capsule to compressive forces exerted on the expanded anterior 
portion of the head dorsum. 

b. Complete division of the eye — The firm contact (along the postocular 
line) of the paraocular and postocular lobes of the parietals apparently re- 
lieves the eye from pressure transferred posteriorly from the head dorsum 
and from antagonistic pressure transferred anteriorly by resistance of the 
prothorax. 

c. Intersegmental inflection of the body wall — Longitudinal compressive 
forces are borne by opposed inflected surfaces which distribute resistant 
forces over much greater areas than would be afforded by the mere opposing 
edges of sclerites. The most extensive inflections occur between the head 
and prothorax and between the prothorax and pterothorax. The narrowed 
posterior portion of the head capsule fits snugly against the prothorax in a 
receptacle formed by the inflection of the anterior prothoracic margin. 
Maintenance of this head-prothorax connection is enhanced in Phanaeus by 
the interlock mechanism formed by the postocular lobes of the parietals and 
emarginations in the anterior portion of the circumnotal ridge. The inflected 
posterior margin of the prothorax is opposed dorsally by anterior inflections 
of the elytra and mesepimera and ventrally by the antero-median promi- 
nence of the metasternum. 

d. Rigidity of posterior part of body — Overall strengthening is achieved 



Evolution of the Phanaeine Dung Beetles 815 

by two complementary devices, firm interlocking and anchoring of the elytra 
and fusion of the abdominal sterna. Elytral interlocking is by means of an 
opposing ridge-groove mechanism along the elytral suture; anchoring of the 
elytra is affected by the metepisternal tabs, which maintain close elytral 
appression along the elytral suture by restricting lateral slippage of the elytra. 
Concealment of all spiracles is apparently another important fossorial 
adaption. The thoracic spiracles are hidden by the overlap of the mesothorax 
by the prothorax; the abdominal spiracles are covered by the elytra. Directly 
exposed spiracles would no doubt be highly susceptible to obstruction by 
soil particles. 

Secondary Modification of Fossorial Features 
As mentioned above, certain basically fossorial characters have equally 
important adaptive significance in other behavioral contexts, such as feeding 
habit. Coprophanaeits is apparently exclusively a necrophagous genus while 
most other phanaeine genera are known to be principally coprophagous. In 
the context of necrophagy, the strong development of clypeal and front tibial 
dentition in Coprophanaeits is especially noteworthy. While dentition in the 
present sense is basically a fossorial adaptation, its degree of development in 
phanaeines is closely related to feeding habit. Strongly developed dentition 
in Coprophanaeits is probably directly related to the principal food of the 
genus, carrion. The resiliency of flesh is obviously much greater than that of 
dung. Hence, structural tools facilitating its collection would presumably 
be selectively advantageous for a necrophagous beetle. I have observed 
Megaphanaeits in Brazil using the foretibiae, employed with unmistakable 
cutting-slicing motions, and clypeus to shear flesh from animal carcasses, in 
the manner suggested by HalfTter and Matthews (1966: 93). That Copro- 
phanaeits is efficient at removing carcasses is well evidenced by the obser- 
vation of Martinez (1959) that four individuals of Coprophanaeits (Mega- 
phanaeits) ensifer cut into pieces and buried a 20 kg (44 lb) dog carcass in 
one night! I observed a somewhat lesser feat performed by five C. (C.) 
jasius, which disposed of an armadillo weighing approximately 2 kg in a 
similar period of time. 

PART V: CLASSIFICATION OF THE PHANAEINES 

That the phanaeine genera form a quite cohesive, natural assemblage has 
been assumed for many years. Early authors referring new taxa to the 
"Phanaei," "group of Phanaeus" etc., recognized the group primarily by the 
form of the antennal lamellae and lack of tarsal claws. These same features 
were stressed by MacLeay (1819) in proposing Phanaeits for certain Linnean 
species of Scarabaeits. 

The first formal taxonomic recognition of a phanaeine assemblage was 



816 The University Science Bulletin 

Kolbe's "unterfam. Phanaeinae" (1905: 550), which included Phanaeus, 
Gromphas, Bolbites, Oruscatus, Megatharsis and Dendropoemon (sic). 
Kolbe apparently overlooked Oxysternon and Diabroctis, both of which were 
proposed for species of Phanaeus long before 1905. Gillet's checklist (1911) 
included the aforementioned genera (except Diabroctis) along with En- 
nearabdus and Pteronyx in Phanaeides, one of three subdivisions (though 
not called subtribes by him) of Coprini. Olsoufiefr (1924) also applied the 
name Phanaeides to essentially the same assemblage as Gillet except that he 
included Diabroctis (as Taurocopris) and Tctramereia (as Boucomontius) 
and excluded (p. 16) both Pteronyx and Ennearabdus. The latter genus has 
since been placed in a tribe of its own, Ennearabdini (Pereira and Martinez, 
1956). Moreover, Olsoufiefr divided Phanaeus into five subgenera, Sulco- 
phanaeus, Coprophanaeus, Metallophanaeus, Megaphanaeus and Phanaeus, 
sen. str. Blackwelder's checklist (1944) follows Olsoufiefr except that the 
above subgenera are each raised to generic rank and Pteronyx and Ennearab- 
dus are readmitted to Phanaeides. 

Most modern workers have adopted a classification of Scarabaeinae based 
upon the keys given by Janssens (1946, 1949), in which the larger tribes are 
subdivided into subtribes. Accordingly, the phanaeine genera comprise 
Phanaeina, one of the three subtribes of Coprini. (The subtribal names of 
Janssens are generally emended to terminate in -ina rather than -ides; thus, 
Phanaeides is now written Phanaeina.) Since Janssens cited neither the 
constituent genera of his subtribes (except those in disagreement with the 
keys) nor bibliography, one can only assume that their taxonomic compo- 
sition agrees with that cited by previous authors. 

For the present I have avoided formal nomenclatural reference to the 
phanaeine genera as Phanaeina or as its nomenclatural equivalents 
(Phanaeini, etc.). Thus, my collective references to this group of genera 
(phanaeines, phanaeine Scarabaeinae, etc.) are intended to be informal 
recognition of a taxon roughly equivalent taxonomically to Phanaeina 
(== Phanaeides = Phanaeini) of authors. My principal reason for at least 
temporarily side-stepping adoption of a formal name is the serious deficiency 
in our understanding of the interrelationships of other genera currently em- 
braced by Coprini as they bear upon the taxonomic position of phanaeines. 
I have, thus, reluctantly concluded that Janssens' classification, at least as 
applicable to Coprini, can at best be used with caution, pending its intensive 
re-examination. Presently it suffices to say that I consider the phanaeine 
genera discussed below a monophyletic taxon characterized by the following 
features : 

I. Antennal club three-segmented; basal lamella more-or-less scoop-shaped 
and enclosing, at least partially, the two distal lamellae; club usually 
large and spherical, but in Dendropaemon often flattened, elongate. 



Evolution of the Phanaeine Dung Beetles 817 

2. Dorsal surface of clypeus at least partially (usually wholly) transversely 
ridged. 

3. Eyes (compound) completely divided into upper and lower portions by 
extension of paraocular areas of the parietals ("genae"). 

4. Front tarsi lacking in males; present or absent in females. 

5. Middle and hind tarsi present, usually five-segmented, claws absent. 

6. Metepisternal tab present. 

7. Medial portion of metasternum acuminate anteriorly, often flattened and 
pentagonal ("rhomboidal"). 

8. Elytral interstriae usually convex, smooth or punctate. 

9. Secondary sexual dimorphism usually strongly pronounced, sexes being 
easily distinguished by configuration of cephalic and pronotal protuber- 
ances, ridges, etc. 

10. Size variable, usually 10-30 mm; typically robust; colors often partially or 
wholly metallic. 

11. Eggs laid singly in earth-covered brood balls (see HalfTter and Matthews, 
1966). 

12. Restricted to New World, mainly neotropical. 

OlsoufiefPs monograph (1924) is still the only comprehensive treatment 
of the phanaeine species. It has, moreover, been the basis of all subsequent 
treatments (including this one). OlsoufiefT evidently had access to most of 
the known phanaeine type material, and his species descriptions and keys to 
species are the standards. There are, however, a number of works postdating 
his which complement it or enlarge upon its reliability as an identification 
manual. Among them are these: Barattini and Saenz (1961, 1964; 
Gromphas), Blackwelder (1944), Blut (1939), Martinez (1944, 1959), 
Martinez and Pereira (1967), Pereira (1949), Pereira and Martinez (1956), 
Pessoa (1934, 1935), Pessoa and Lane (1936, 1941), Robinson (1947) and 
Vulcano and Pereira (1967:565 ft). The supraspecific classification presented 
here differs from that of OlsoufiefT in the following respects: 

1. Splitting of Phanaetts into three genera. The following outline shows 
the general changes, details of which are brought out below : 

Olsoufieff (1924) Present 

Genus Phanaeus 

Subg. Sulcophanaeus Genus Sidcophanaeus 

Subg. Phanaeus, sen. str. Genus Phanaeus 

Genus Coprophanaeus 
Subg. Megaphanaeus Subg. Megaphanaeus 

Subg. Metallophanacus Subg. Metallophanaeus 

Subg. Coprophanaeus Subg. Coprophanaeus, sen. str. 



818 The University Science Bulletin 

2. Division of Oxystemon and Dendropaemon into two and three sub- 
genera, respectively. 

3. Recognition, where practical, of species-groups within generic cate- 
gories. (This was attempted formally by OlsoufiefT only for Phanaeus, 
sen.str.) 

4. Removal of Bolbites, Gromphas and Oruscatus from the phanaeine 
assemblage. 

Removal of Bolbites, Gromphas and Oruscatus from "Phanaeina" is done 
confidently but, at the same time, reluctantly since their alternative positions 
in the current classification of Coprini are not in the least clear. The rather 
strict definition of phanaeines given above definitely excludes Gromphas and 
Oruscatus, at least morphologically. Both are probably more closely related 
to dichotomies or, perhaps, to Old World elements which I have not 
studied. Nevertheless, as far as the South American fauna is concerned, both 
are taxonomic isolates, closely related neither to each other nor to other ex- 
tant groups with which I am familiar. Bolbites onitoides Har. (monobasic), 
on the other hand, may later prove to be a phanaeine in some wider taxo- 
nomic sense than that presently used. It appears to show features of both 
phanaeines (some aspects of head structure, general body shape, metepisternal 
tab, etc.) and nonphanaeines (presence of tarsal claws, antennal club not of 
Phanaeus-type, etc.). It could very well happen that this species will prove 
to be highly important in assessing the position of phanaeines relative to the 
rest of the tribe. Nevertheless, firm placement of these three genera must 
await a thorough re-examination of Coprini on a world-wide basis. 

The following classification is based upon phenetics and, therefore, sub- 
ject to an inherent degree of inadequacy. The implicit hope is, of course, 
that the characters used, although a small sample of those potentially available, 
have been selected and studied sufficiently well that results based upon them 
will be supported by newly discovered characters. The terminology used in 
generic descriptions, keys, etc., is that established above; in doubtful in- 
stances, readers should refer to the two preceding sections. Some character 
states describing genera are numbered as are their definitions in part III; 
salient features are italicized in generic descriptions. 

A potential source of inconvenience is that I was able to examine very 
little type material. The result could be perpetuation of a certain amount of 
nomenclatural confusion attending the names of some of the more poorly 
known species. I suspect that the greatest potential for error lies in the genus 
Dendropaemon, most of the species of which are only poorly known and 
subject to conflicting descriptions and usage in the literature. Nevertheless, 
on the whole I do not anticipate a great deal of subsequent need for cor- 
rection since the majority of species names have been used quite consistently 
for many years. Moreover, during the course of this study, I consulted vari- 



Evolution of the Phanaeine Dung Beetles 



819 



ous workers who together have examined a great deal of the available type 
material; among them is Padre F. S. Pereira of Sao Paulo. 

Below is an outline of the classification proposed in the remainder of this 
part: 



Genus DIABROCT1S Gistel 
Genus SULCOPHANAEUS Olsoufieff 
faunus-group 
imperator-group 
auricollis-growp 
Genus PHANAEUS MacLeay 
splendidulus-gxoup 
palaeno-complex 
endym ion-complex 
chalco m elas-com plex 
splendidnlus-complex 
bispinus Bates (an isolate) 
vindex-group 

f index-complex 

mexicanus-complex (comprising 
several species-clusters) 
Genus OXYSTERNON Castlenau 
Subgenus OXYSTERNON, sen. str. 
jestivum-group 
sericen m -grou p 
Subgenus MIOXYSTERNON 

NEW SUBCENUS 



Genus COPROPHANAEUS Olsoufieff 
Subgenus MEGAPHANAEUS 
Olsoufieff 
Subgenus M ET A LLO PHANAEUS 
Olsoufieff 

Subgenus COPROPHANAEUS, 
sen. str. 

jasius-group 
dardanus-group 
Genus DENDROPAEMON Perty 
Subgenus PARADENDRO- 

PAEMON NEW SUBGENUS 

Subgenus COPROPHANAE- 

OIDES NEW SUBGENUS 

Subgenus DENDROPAEMON, 
sen. str. 

dentico/lis-group 
viridipennis-groxip 
piceum-group 
Genus TETRAMERE1A Klages 
Genus HOMALOTARSUS Janssens 
Genus MEG ATH ARSIS Waterhouse 



Artificial Key to the Genera of Phanaeine Scarabaeinae 

The following key is meant principally for purposes of identification. An 
effort has been made to select those character states most easily seen. Infrageneric 
taxa are separated in discussions of genera. 

1. Middle and hind tarsi each with fewer than five segments; front tarsi 

always absent in female 2 

-Middle and hind tarsi five-segmented; front tarsi present or absent in 

female 3 

2. Middle and hind tarsi four-segmented, fourth (distal) segment very 

small (Fig. 263); clypeus not emarginate medially (Fig. 95); body 

robust (Figs. 312-313), color red-brown Tetramereia Klages 

-Middle and hind tarsi either two or three-segmented (Figs. 270-279); 
clypeus emarginate medially (Figs. 89-94); body robust to flattened 
(Figs. 304-309); color usually at least partially metallic blue or 
green Dendropaemon Perty 

3. Dorsal surfaces of hind tibiae and basal four hind tarsal segments cov- 

ered by large, setigerous granules (Fig. 260); tarsal segments as in 
Figures 264 and 265; color dull metallic green; body robust (Figs. 
310-311) Homalotarsus Janssens 



820 The University Science Bulletin 

- Dorsal surfaces of hind tibiae and tarsi glabrous or with small 

setigerous granules; body robust, color variable 4 

4. Labial palpus comprising a single oval segment (Figs. 154-155); dorsal 

surface of hind tibiae (Fig. 266) and abdominal sterna (Fig. 301) 

bearing dense fringes of rather long setae Megatharsis Waterhouse 

-Labial palpi three-segmented; tibiae and abdominal sterna without 

dense fringes 5 

5. Anterior margin of clypeus emarginate medially (Figs. 83-89), clypeal 

process U-shaped; front tarsi usually absent in female; front tibiae 

as in Figure 246 Coprophanaeus Olsoufieff 

-Anterior margin of clypeus not emarginate medially (Figs. 98-110), 
clypeal process never U-shaped; front tarsi rarely absent in female; 
front tibiae variable (Figs. 250-259) but not as in Figure 246 6 

6. Transverse clypeal carina present (Figs. 194, 197); transverse hind 

tibial carina present; large, green species Diabroctis Gistel 

-Transverse clypeal carina absent; hind tibial carina rarely present; size 

and color variable 7 

7. Antero-median angle of metasternum prolonged as a long, sharp spine 

extending between apices of the front coxae (Figs. 244-245); pos- 
tero-median angle of pronotum acute, distinctly prolonged between 

bases of elytra (Fig. 210) Oxy station Castlenau 

-Antero-median angle of metasternum not spinate; postero-median 

angle of pronotum not distinctly prolonged 8 

8. Occipital ridge absent or incomplete; anterior portion of circumnotal 

margin not broken by emarginations behind eyes (Figs. 117C, 12 1C- 

122C) Sulcophanaeus Olsoufieff 

-Occipital ridge complete (Figs. 1, 4); anterior margin of circumnotal 
ridge distinctly interrupted by emarginations receiving postocular 
lobes of parietals (Figs. 35, emarg; 36-38) Phanaeus MacLeay 

Genus Diabroctis Gistel 

Diabroctis Gistel, 1857: 604. Type: Scarabaeus mimas L., original designation. 
Tattrocopris Olsoufieff, 1924:18. Type: Scarabaeus mimas L., present designation. 

Until recently this genus has been known as Tattrocopris Olsoufieff; the 
generic synonymy, Taurocopris = Diabroctis, was established by Martinez 
and Pereira (1967). Olsoufieff designated no type-species for Tattrocopris. 
In order to stabilize the use of these names by precluding future question 
about their synonymy, it seems appropriate to hereby designate as the type- 
species of Tattrocopris Olsoufieff (1924) the following species: Scarabaeus 
mimas L. By so doing, Tattrocopris automatically becomes a junior objective 
synonym of Diabroctis Gistel. 

DESCRIPTION 

(1) anterior margin of clypeus not emarginate medially, strongly toothed 
(Figs. 98-99); (2) clypeal process transverse; lateral clypeal carinae well 
developed; transverse clypeal carina present (Figs. 194, 197, tcrcly); (3) 



Evolution of the Phanaeine Dung Beetles 821 

cephalic process of male lamellate, a transverse ridge prolonged on each side 
into acute "horn" (Figs. 194-195) (except mirabilis where lateral prominences 
greatly prolonged); (4) cephalic process of female carinate (Fig. 197); (5) 
paraocular areas of parietals narrow; (6) postocular lobes of parietals de- 
pressed transversely (Fig. 115B), anterior portion circumnotal margin not 
emarginate (115C); canthal area distinct (Fig. 115A, ca) ; (7) cephalic 
brachium monopodal; (8) eyes small; (9) occipital ridge incomplete, repre- 
sented only by lateral segments; (10) labrum Phanae us-type; (11) mentum 
strongly emarginate; (12) median lobe of hypopharynx present; (13) pre- 
mental sclerites marginally sclerotized; (14) pronotum of male type C (Figs. 
194-195); (15) pronotum of female type D (Fig. 196); (16) pronotum 
punctate (rugose antero-laterally in mimas) ; posterior pronotal fossae present 
(except mirabilis), large; (17) front tarsi present in female; (18) front tibiae 
type C (Fig. 250); hind tibiae carinate laterally (as in Fig. 262); (19) mid- 
dle and hind tarsi five-segmented; (20) elytral interstriae smooth and (21) 
evenly convex; (22) striae simple; hind wing not notched; color dar\, dull 
green to metallic yellow-green; size large (length 25-40mm). 

The three species included in Diabroctis {mimas [L.], cadmus [Har.], 
mirabilis |Har.]) are easily recognized by their green color, large size and a 
transverse clypeal carina anterior to the cephalic process. D. mimas and 
cadmus are among the largest phanaeines (40 mm) and appear to be closely 
related; the male prothorax of these species (Figs. 194-195) is exceptionally 
massive. D. mirabilis differs from its congeners in the features brought out 
above, the most striking of which is the massive, antler-like development of 
the male cephalic process. All species have a distinct fringe of setae projecting 
from the elytral suture (the setae insert on the elytral margins), which is 
very prominent in mirabilis. Some species of Phanaeus (e.g., demon') have 
such fringes. 

Diabroctis is restricted to South America; although cadmus has been 
cited (erroneously?) from Costa Rica (Gillet, 1911; Black welder, 1944), it 
is probably only South American (Colombia). D. mimas occupies an enor- 
mous range essentially determining the distributional limits of the genus 
while mirabilis is common only in southern Brazil and adjacent regions. A 
map of the range of Diabroctis is not given since it would look essentially 
like that given for Oxysternon, sen. str. (Map 3). Little is known about the 
biology of Diabroctis but it appears basically coprophagous. Both mimas and 
mirabilis have been collected in large numbers from human and bovine feces; 
mimas is often collected in swine feces in Manaus, Amazonas (Brazil). 

Genus Sulcophanaeus Olsoufieff 

Sulcophanacas Olsoufieff, 1924: 23 (as subgenus of Phanaeus). T\pe: Scarabaetis sulcatus 
Drury = Scarabaetis carnijex L., 1758, not 1767, original designation. 

Olsoufieff (1924: 23) designated Phanaeus sulcatus (Drury) as the type- 



822 The University Science Bulletin 

species of the subgenus Sulcophanaeus. The name sulcatits Drury has since 
been recognized as a junior subjective synonym of carnijex L., 1758 (Mat- 
thews, 1966: 39). Consequently, the valid name of the type-species of this 
genus becomes Phanaeus carnijex (L.), 1758. It should be noted that, pend- 
ing a ruling by the International Commission on Zoological Nomenclature, 
the name Sulcophanaeus could later be ruled a junior objective synonym of 
Onthurgus Gistel (see discussion below in introduction to Phanaeus). 

description 

(1) anterior margin of clypeus not emarginate medially (Figs. 102-105); 
(2) clypeal process transverse, either tooth-like or ridge-like; lateral clypeal 
carinae usually present; (3) cephalic process of male corniform (Figs. 198, 
203-204, 209); (4) cephalic process of female corniform or carinate (Figs. 
199, 201-202, 205, 207); (5) paraocular areas of parietals narrow; (6) post- 
ocular lobes of parietals depressed transversely (part B of Figs. 117, 121-122), 
circumnotal ridge not interrupted by emarginations (parts C of same Figs.) ; 
canthal area usually distinct (Fig. 117); (7) cephalic brachium monopodal 
(bipodal in few species); (8) eyes small; (9) occipital ridge incomplete; 
(10) labrum Phanaeus-typc; (11) mentum strongly emarginate (Fig. 158); 
(12) median lobe of hypopharynx present; (13) premental sclerite marginally 
sclerotized (Fig. 150); (14) pronotum of male variable (Figs. 198, 203-204, 
206, 209); (15) pronotum of female type C or D (otherwise in faunus 
group, Figs. 199-200); (16) pronotum smooth (rugose in laeander and 
columbi); anterolateral angles of pronotal ridge often angulate; posterior 
pronotal fossae present (except actaeon)\ (17) front tarsi present in female 
(except carnijex and, reportedly, steinheili)\ (18) front tibiae Phanaeus-typc; 
transverse hind tibial carinae sometimes present; (19) middle and hind tarsi 
five-segmented; (20) elytral interstriae smooth and (21) flat or convex; (22) 
striae simple; hind wing not notched. 

The name Sulcophanaeus is used here in a much different sense from 
that used by Olsoufiefr (1924). In the subgenus Sulcophanaeus Olsoufiefr 
included only two species, carnijex (as sulcatits) and stcinheili, which re- 
portedly resembled each other in the structure of the male genital capsule 
and lack of front tarsi in the female; but he was doubtful about the sound- 
ness of his grouping. In the present sense Sulcophanaeus brings together a 
rather diverse group of species including not only the above two but also a 
large number included by Olsoufiefr in Phanaeus, sen. str. (all of species 
group 1, species group 2 [except bispinus) and part of species group 3). The 
genus forms three species groups which, because of their distinctness, would 
likely warrant consideration as subgenera if the genus were larger. While 
these species groups are certainly distinct, it is noteworthy that each, inter 



Evolution of the Phanaeike Dung Beetles 823 

se, comprises more or less readily recognizable isolated elements (phenetically 
and often geographically) mostly as single species or species pairs. 11 

Key to the Species Groups of Sulcophanaeus 

1. Both sexes with well developed corniform cephalic processes, that of 

the male very long, curved, tapering (Fig. 198), that of the female 
either similar to male {jaunus, Fig. 199, and columbi), strongly 
bifurcate (laeander) or as in Figures 201 and 202 (carnijex); para- 
ocular areas of parietals each strongly carinate above lateral to eye; 
postocular lobes of parietals, seen laterally, extended anteriorly as 
finger-like process (Fig. 117A) approaching upper portion of eye 
(except carnijex); clypeal process tooth-like (Fig. 113); lateral 
pronotal carina present (Figs. 199, 201, 203, lpc); elytral interstriae 
unevenly convex ("sulcate") (Fig. 283); large beetles, either uni- 
form dull black brown or metallic green jaunus group 

-Cephalic process of male as above; that of female a simple or tuber- 
culate carina (Figs. 205, 207); paraocular areas carinate or not; 
postocular lobes of parietals not projecting anteriorly; clypeal 
process ridge-like (except rhadamantus); lateral pronotal carina 
absent (except male imperator); elytral interstriae convex or flat; 
moderate sized (20-30mm), color variable 2 

2. Paraocular areas of parietals not strongly carinate lateral to eye; front 

tibiae tridentate, teeth concentrated in apical 1/3 of outer margin 
(Fig. 251); elytral interstriae flat (Fig. 282), though sometimes 
with irregular wrinkling; pronotum of male flat or shallowly con- 
cave, sloping downward anteriorly from pair of closely set postero- 
median processes (type F); pronotum of female convex except for 
anterior transverse ridge shaped as in Figure 207; elytra dull black 
(except actaeon), pronotum completely or marginally metallic. 

auricollis group 

-Paraocular areas of parietals carinate lateral to eye; apical three teeth 
of front tibiae occupying more than 1/3 outer margin; elytral inter- 
striae convex; pronotum of male variable; pronotum of female 
variable; color variable imperator group 

the jaunus group (f annus |F.|; columbi [MacL.]; laeander [Water- 
house | ; carnijex [L.|): All these species are large insects (30 + mm) and, in 
addition to the features mentioned in the key, resemble each other quite 
closely in general appearance of the head capsule. Although they are easily 
distinguishable, columbi and laeander form a loose species pair (rugose 
pronotum, metallic green color, bipodal cephalic brachium). Both, however, 
are clearly related to jaunus (see Key), which at the same time shares a 
number of unique features with carnijex (strong transverse hind tibial 
carinae, uniform dull black-brown color, intrasternal position of last ab- 



11 In this and subsequent discussions a species pair is meant to refer to two species which 
are presumed more closely related to each other than either is to other members of a given 
taxon. In general the members of a pair are very closely related and sometimes difficult to 
separate (i.e., "sister species"), but this need not be so. 



824 The University Science Bulletin 

dominal spiracle). S. jaitnus, then, is an annectant species connecting other- 
wise quite different species. The general resemblance of fannus and carnijex 
is striking and was mentioned, but not stressed by Olsoufieff (p. 79). 

S. carnijex is noteworthy in that, being endemic to Jamaica, it is the only 
phanaeine not a continental species. The most reasonable explanation for its 
presence there is that, like a number of other Jamaican scarabaeines (Mat- 
thews, 1966), it is an insular relict of probably very long standing residence. 
The relict status of carnijex as well as its ecology is discussed by Matthews 
(1966). The other species in this group are South American (see Map 1). 
The ranges of laeander and columbi are poorly known due to the rarity of 
these species in collections; the former is generally considered Venezuelan 
(a savanna species ?) while columbi has been collected in the southern part 
of the Amazon Basin. S. jaitnus, which is definitely a forest species, occupies 
a wide range in the Amazon Basin. Except for carnijex, the biology of these 
species is largely unknown. Matthews (1966) reports that carnijex is copro- 
phagous, occurring in large numbers in cow dung, and likely restricted to 
certain soil types. This is the only instance where the possible limiting effects 
of edaphic factors upon phanaeines has been reported; the distributional 
effects of soil types on Scarabaeinae has so far been largely ignored. S. 
columbi has been collected from tapir feces (B. Malkin, pers. comm.) while 
jaunus is reportedly necrophagous( Pereira, pers. comm.). I know nothing 
of the biology of laeander, but if it truly is a savanna species, I would expect 
it to be coprophagous. 

the imperator group {imperator [Chevr.|; stein heili [Har.]; rhadaman- 
tus [Har. j; batesi [Har.]; menelas [Cast.]): This group consists of two 
species pairs and one species which appears taxonomically isolated. The 
latter species, rhadamantus, is extremely rare; only because it somewhat re- 
sembles imperator do I place it here. S. batesi and menelas (splendididus of 
Olsoufieff and often so identified in collections) are very similar (unique 
shape of pronotum of both sexes, "sulcate" elytra), except for color differing 
only in minor details. A study of the geographic variation of this pair is in 
order since they could be subspecies. A rather close relationship between 
imperator and steinheili can be supported here only on the basis of the males 
(very similar pronota; presence of hind tibial carinae). This similarity was 
mentioned both by Harold (1875) and Olsoufieff (p. 79). I have not seen 
the female of steinheili, which, according to Olsoufieff, lacks front tarsi. This 
species pair only slightly resembles menelas-batesi in general facies. The 
presence of hind tibial carinae and lateral pronotal carinae link it with 
jaunus-carnijex, which it resembles even less in general appearance. More- 
over, imperator has a bipodal cephalic brachium, known elsewhere in the 
genus only in laeander-columbi. Such a mosaic appearance of features can 



Evolution of the Phanaeine Dung Beetles 



825 



lY, 



*5*£ 



iX 



-o 



MAP 1 



Sulcophanaeus 
SJgs? carnifex group 
:::$:$ imperator group 
| auricol I is group 



V 

\ 




J%? 



Map 1. Approximate distributions of the species groups of Sulcophanaeus. Arrows connect 
widely disjunct portions of distributions. 

be interpreted as either parallel evolution or a closer phyletic proximity than 
indicated by the classification; I have chosen the former interpretation. 

The imperator group comprises largely temperate species. Except for 
steinheili, a rare species known only from the Bogota region of Colombia, 
they occur in southern South America (Map 1). S. imperator, the southern 
relative of this species, is also a high altitude species inhabiting the higher 
Andean plateau regions of Argentina and Bolivia. The range of batesi 
approximates that of imperator while menelas occurs in the grassland regions 
of eastern and central Argentina and Bolivia and southern Brazil. S. rhada- 
mantus is known from very few specimens, some from the state of Buenos 



826 The University Science Bulletin 

Aires, Argentina. Martinez (1959), reports that imperator, batesi and 
menelas are strictly coprophagous, preferring herbivore (livestock) excre- 
ment. The latter species, however, is evidently at least occasionally necro- 
phagous (J. Foerster, pers. comm.). 

the auricollis group {auricollis [Har.]; chryseicollis [Har.]; noctis 
[Bates]; velutinus [Murray]; actaeon [Erichson]): Of the species groups 
comprising Sukophanaeits this is the most cohesive, phenetically and geo- 
graphically. Three of the five included species, auricollis, chryseicollis and 
noctis, appear quite closely related, differing most obviously in color. While 
the pronotum of the male of velutinus is strikingly different from that of the 
above three species (cf., Figs. 206, 209), the pronotum of the females of 
all four are very similar, principally because of the unique shape of the 
anterior carina (Fig. 207). S. actaeon, the female of which I have not seen, 
differs from the above species in several respects, the most striking of which 
is its all metallic color. While the elytral interstriae are flat and dull black 
in other species, in actaeon they are irregularly raised (appearing "blistered") 
and metallic blue or green. To a much lesser degree, the "blistering" is seen 
in other species of the group, particularly velutinus. A curious feature of this 
group, seen also in some Phanaeus, is a tendency for the propleura to be 
hollowed for the reception of the posterior margins of the paraocular areas 
of the parietals. The depth of these concavities varies as does the size of the 
tubercle adjacent to each which fits beneath the paraocular area. 

The center of distribution of this group appears to be the lowlands and 
foothills along the eastern slopes of the Andes from Panama southward 
through Peru (Map 1). One species, chryseicollis, is largely restricted to 
neotropical southern Mexico and is the only continental Sulcophanaeus not 
occurring also in South America. Both noctis and velutinus occur from 
Panama to Peru; auricollis is largely restricted to Colombia while actaeon is 
known to me from only a few Peruvian localities. None of these species, 
however, are well represented in recent collections with precise locality 
designation. Accordingly, in view of the tremendous physiographic variation 
of the general range of this group, very little is known of their ecological 
distribution within it. I have no ecological data on actaeon and auricollis. 
S. chryseicollis has been collected in southern Mexico at night using traps 
baited with human excrement (Halffter and Matthews, 1966). Both noctis 
and velutinus have been collected in the same manner from the Chiriqui 
region of Panama (D. Viers, pers. comm.). 

Genus Phanaeus MacLeay 

Phanaeus MacLeay, 1819: 124. Type: Scarabaeus camijex L., 1767, not 1758 = Phanaeus 
vindex MacLeay, subsequent designation by Olsoufieff, 1924: 23, see discussion below. 

Lonchophorus Germar, 1824: 126. Type: Scarabaeus camijex L., 1767, not 1758 = Phanaeus 
vindex MacLeay, present designation. 



Evolution of the Phanaeine Dung Beetles 827 

Onthurgus Gistel, 1857: 602. Type: Scarabaetis carnifex L., 1767, not 1758 = P/ianaeus vindex 
MacLeay, original designation (objective generic synonymy subject to opinion of the 
International Commission on Zoological Nomenclature, see below). 

Phanaeus, sen. str., OlsoufiefT, 1924: 23 (pars). 

The following nomenclatural melodrama has emerged in determining 
:he valid name of the type-species of Phanaeus. MacLeay (1819) designated 
no type-species for Phanaeus, nor were any of the included species indicated 
as type-species. Subsequently, OlsoufiefT (1924: 23) designated as the type- 
species of Phanaeus, sen. str., a North American species which he referred to 
as P. carnifex (L.). Although he rightly suspected that Linnaeus had used 
the name carnifex in at least two different senses (one in 1758, another in 
1767), OlsoufiefT nevertheless synonymized the names carnifex L. and vindex 
MacLeay (pp. 95 and 151). Matthews (1966: 41), in order to rectify the 
confusion attending the name carnifex, designated a lectotype fixing the 
name to a Jamaican species and thereby preserved Linnaeus's original usage 
of the name carnifex (1758). Consequently, carnifex becomes the valid name 
of the type-species of Sulcophanaeus (see above), and vindex, the only 
available synonym for carnifex sensu Linnaeus (1767) and OlsoufiefT (1924), 
becomes the valid name of the type-species of Phanaeus. The following 
synonymies are based on several sources : 

Sulcophanaeus carnijex (L.) 

= Scarabaeus carnifex L., 1758: 346 (not 1767) 

= S. siilcatus Drury, 1770 

= S. sphinx Fab., 1772 

= S. belzebul Fab., 1775 

= S. nicanor Fab., 1781 

= S. /arHerbst, 1789 

= Phanaeus moloch MacLeay, 1819: 129 (species inquirenda) 

= P. siilcatus (Drury), OlsoufiefT, 1924: 23 
Phanaeus vindex MacLeay 

zz Scarabaeus carnijex L., 1767: 546 (not 1758) (junior primary homonym) 

= Phanaeus vindex MacLeay, 1819: 133 (species inquirenda) 

— P. vindex MacLeay, 1833: 47 (from Gillet, 1911: 82) 

= P. carnifex (L.), Gistel, 1857: 602 

= P. carnifex (L.), OlsoufiefT, 1924: 23, 39, 95, 151 

The mistaken identity of carnifex L., 1767, by Linnaeus himself and 
subsequently by OlsoufiefT, has resulted in a case of mis-identified type-species 
of a genus. In accordance with Articles 67 (j) and 70 (a) of the International 
Code of Zoological Nomenclature, I have requested that the International 
Commission on Zoological Nomenclature rule Scarabaeus carnifex L., 1767, 
not 1758, = Phanaeus vindex MacLeay the type -species of Phanaeus Mac- 
Leay. The use of the name carnifex by OlsoufiefT is unquestionably in the 
sense of Linnaeus, 1767, hence in the sense of vindex MacLeay (see Olsou- 
fiefT, 1924: 95). 

As for the name Onthurgus Gistel, the situation is not quite so simple. 



828 The University Science Bulletin 

Gistel (1857: 602) designated as the type -species of this genus "Phanaeus 
camifex (L.)", leaving uncertain in which sense of Linnaeus (1758 or 1767) 
the name was used. The Code states (Art. 70) that one should assume that 
names designated as type-species of genera are used correctly unless there is 
evidence to suggest otherwise. To assume correct usage in this case would 
be to assume that Gistel used the name carnijex in the sense of Linnaeus, 
1758, thereby requiring that Onthurgus be regarded a senior objective 
synonym of Sulcophanaeus. Since there are several reasons favoring the re- 
tention of the name Sulcophanaeus (particularly the taxonomic relationships 
implicit in the "phanaeus" part of the name), I have requested that the Com- 
mission 1) exercise its plenary powers to suspend Article 70 of the Code and 
rule that the type-species of Onthurgus Gistel is Scarabaeus carnijex L., 1767, 
not 1758, = Phanaeus vindex MacLeay and thereby effect the objective syn- 
onymy of Onthurgus and Phanaeus and/or 2) rule the name Onthurgus 
Gistel a nomen oblitum in accordance with Article 23 (b) of the Code since, 
to my knowledge, the name has never been used since 1857. A positive ruling 
by the Commission on either (1) or (2) will preserve the present usage of 
the name Sulcophanaeus. 

DESCRIPTION 

(1) anterior margin of clypeus not emarginate medially, usually toothed 
(Figs. 4, 105-110); lateral clypeal carinae almost always present; (2) clypeal 
process transverse (rarely spiniform); (3) cephalic process of male corni- 
form (Figs. 1, 216, 218, 221, 226); (4) cephalic process of female carinate 
(Figs. 3, 79, 81-82, 220, 222); (5) paraocular areas of parietals narrow; (6) 
postocular lobes of parietals depressed obliquely (Fig. 4), prominent, fitting 
in emarginations of circumnotal ridge (Fig. 35, emarg; not distinct in 
haroldi, Fig. 119); canthal areas not distinct; (7) cephalic brachium bipodal 
(anterior arm reduced in some species, Figs. 129-130); (8) eyes small; (9) 
occipital ridge complete (Figs. 1, 4, 8, ocr) ; (10) labrum Phanaeus-type 
(Figs. 18-19); (11) mentum strongly emarginate (Fig. 25); (12) median 
lobe of hypopharynx present; (13) premental sclerites completely sclero- 
sed; (14) pronotum of male type A or B (Figs. 35-36, 216-219, 221, 223- 
228); (15) pronotum of female type B or D (Figs. 37-38, 220, 227); (16) 
pronotal sculpturing punctate or roughened (Figs. 159, 161, 163-165) ; pos- 
terior pronotal fossae usually present; (17) front tarsi present in female 
(Figs. 61, 63); (IS) front tibiae tridentate or quadridentate (Figs. 60-61, 
256-259); (19) middle and hind tarsi five-segmented (Figs. 64-66); (20) 
elytral interstriae smooth (Figs. 285-287) or rugose (Fig. 284); (21) inter- 
striae flat or evenly convex; (22) striae usually simple; hind wing not 
notched (except bispinus); last abdominal spiracle marginal; color only 
rarely not at least partially metallic. 

The name Phanaeus is used here in a more restricted sense than has 



Evolution of the Phanaeine Dung Beetles 829 

heretofore been customary. It is roughly equivalent taxonomically to species 
groups 3 and 4 of the subgenus Phanaeus, sensn Olsoufieff (1924). Taxo- 
nomically Phanaeus is the most difficult phanaeine genus because of the high 
incidence of irregular distribution of character states and of continuous vari- 
ation of characters reliable (i.e., with discrete states) in other genera. Thus, 
there exist in the genus (particularly evident in the mexicanus complex) 
extremes which definitely warrant taxonomic recognition; but, on the 
other hand, separation of the extremes is precluded by an array of inter- 
mediate forms. In the interest of a balanced classification, I have concluded 
it advisable to recognize only informally what appear to be the constituent 
taxa of this genus. I have, therefore, used freely the following categories: 
species-group, species-complex and species-cluster. Further study using more 
characters will be necessary before the applicability of more formal categories 
(e.g., subgenus) can be determined. 

Key to the Species Groups and Species Complexes of Phanaeus 

1. Pronotum evenly, strongly roughened (granulate to rugose) at least 

antero-laterally but often entirely so; punctures, it present, re- 
stricted postero-medially; cephalic process of female an isolated 
carina or thick ridge (Fig. 82) (except homes cluster); pronotum 
of male with prominent postero-lateral angles, usually appearing 

more-or-less triangular from above vindex group 6 

- Pronotum completely or largely punctate, often appearing glassy 
smooth and shining, at most with irregular gibbosities (except fe- 
male of meleagris) or minutely granulate (males of endymion com- 
plex); cephalic process of female a transverse, usually trituberculate 
carina extending fully between mesal ends of lateral clypeal carinae 
(Fig. 81); pronotum of male variable splendidulus group 2 

2. Pronotum of male with close-set pair of parallel, elongate processes (as 

in Figs. 214-215); that of female with a single such process project- 
ing forward from middle of posterior margin of an antero-median 
concavity; elytral striae carinulate, distinctly punctured; hind wing 

deeply notched; dorsum dull greenish-brown bispinus Bates 

-Pronotum of male as in Figures 216-219, 221; that of female as in Fig- 
ures 220, 222; elytral striae variable; hind wing not notched; color 
variable 3 

3. Clypeal process small, spinate; cephalic process of small males a short, 

bituberculate process; elytral striae carinulate, not punctate; dorsum 
dark, highly shining green; anterior margin of clypeus strongly 
toothed medially; venter often bearing long, whitish hair .. palaeno complex 
-Clypeal process variable, not spinate; cephalic process of small males 

never a bituberculate process; other characters variable 4 

4. Front tibiae of female quadridentate (Fig. 256), resembling Copro- 

phanaeus-type; anterior margin of clypeus strongly toothed medi- 
ally (Fig. 108); pronotum of male very flat dorsally, produced 
postero-laterally and appearing triangular from above (Figs. 218- 
219), sometimes very finely and irregularly granulate dorsally but 



830 The University Science Bulletin 

usually appearing smooth without magnification; pronotum of fe- 
male smooth, without granulation endymion complex 

-Front tibiae of female tridentate; anterior margin of clypeus at most 

only weakly toothed medially; other characters variable 5 

5. Elytral striae carinulate, distinctly punctate; pronotum bearing irregu- 

larly spaced gibbosities dorsally, punctate antero-laterally (more-or- 
less evenly, though not strongly rugose in meleagris); clypeal 
process not tooth-like; pronotum of male very flat dorsally, pro- 
duced postero-laterally and appearing triangular from above; 
dorsum yellow-brown ("bronze ) or nearly black .... chalcomelas complex 
-Elytral striae simple, not punctate; pronotum punctate or smooth; 
clypeal process tooth-like (as in Fig. 113); pronotum of male deeply 
concave dorsally (Fig. 217); color variable splendidulus complex 

6. Elytral interstriae rugose (Fig. 284) or (seen magnified) distinctly 

heavily punctured, punctures coalescent vindex complex 

-Elytral interstriae smooth or only very finely punctured (Figs. 285- 

286) mexicanus complex 

The splendidulus group: This is essentially a South American group 
(Map 2) the members of which fall rather neatly into the species complexes 
separated above. The palaeno complex is a species pair (palaeno Blanch, and 
kjrbyi Vigors) ; palaeno bears a striking resemblance to Oxysternon palae- 
mon, the two species forming the only potential intergeneric mimic-model 
complex among phanaeines. Both species of this complex occur (sym- 
patrically ?) in the savanna-turned-pasturelands of southern Brazil and ad- 
jacent regions; palaeno (and also \irbyi ■) is very common in cattle feces. 

The endymion complex includes three closely related species (endymion 
Har.; pyrois Bates; blanchardi Ols.) with quadridentate front tibiae re- 
sembling those of Coprophanaeus. Both endymion and pyrois occur in 
tropical lowland forests of Central America and southern Mexico and are 
the only yet described members of the splendidulus group occurring outside 
mainland South America. These two species have been collected from traps 
baited with human feces. I know blanchardi from a single locality in Ecua- 
dor ("Pucay"). There is a fourth, yet undescribed species likely belonging 
in this complex which I know from a single, well developed male collected 
near Temascaltepec, Mexico (state of Mexico) by H. E. Hinton. The male 
in question is evidently one of several individuals collected by Hinton from 
a decomposing hymenomycete fungus (probably a toadstool) (Hinton, 
1935). Although referred to by him as endymion, which the new species 
closely resembles, Hinton's illustration is undoubtedly of a well developed 
male of the same new species, not endymion. (P. funereus Balths. is an 
unassigned name probably belonging in this complex.) 

The splendidulus complex includes three rather closely related species 
(splendidulus [Fab.], dejeani Har. and melibaeus Blanch.) in addition to 
haroldi Kirsch. The first two species have been collected from cow dung in 



Evolution of the Phanaeine Dung Beetles 



831 



MAP 2 




Phanaeus 

vindex group 

vindex complex ' 

mexicanus complex ■; 

splendidulus group -z< 



Map 2. Approximate distributions of the species groups of Phanaeus. 



the Serra do Mar region of southern Brazil hut neither appears restricted to 
the area. Some confusion exists with locality data under the name splendi- 
dulus since until recently the name was erroneously applied to Sulcophanaeus 
menelas. P. melibaeus is restricted to the southern Amazon Basin where it 
has been collected in dung. The shape of the pronotum and aspects of head 
structure set haroldi apart from other members of the complex; I have in- 



832 The University Science Bulletin 

eluded it here, perhaps artificially, on the basis of its closer resemblance to 
melibaeus than to any other species in the group. The range of haroldi 
appears concentrated in the tropical lowland valleys along the eastern slopes 
of the Andes (Colombia, Ecuador and Peru) . 

The chalcomelas complex is a species pair (chalcomelas Perty and 
meleagris Blanch.). The former is a widespread Amazon Basin species 
which is reportedly coprophagous (Halffter and Matthews, 1966: 31). It and 
meleagris, a necrophagous species (he. eit., 31), bear distinct, though usually 
irregular gibbosities over the surface of an otherwise punctate pronotum 
which suggest the pronotal roughening of the vindex group. 

Phanaeus bispinus Bates (an isolate of complex rank) resembles both 
chalcomelas (carinulate-punctate elytral striae, color) and haroldi (shape of 
cephalic process and other aspects of head structure), but several unique 
features support its taxonomic isolation with the splendidulus group: shape 
of the pronotum of the male (not unlike that of Oxysternon curvispinum, 
Figs. 214-215) and of the female (which is unlike any other phanaeine, a 
single elongate process in place of the paired processes of the male) and 
notching of the hind wing (seen otherwise only in some Coprophanaeus 
and in Dendropaemon). I have seen few specimens of bispinus, all from the 
Amazon Basin; Pereira and Martinez (1956) say it is necrophagous. 

The vindex group: This species group includes the bulk of those species 
here assigned to Phanaeus (32 of 44 species). Most are North American 
species (Map 2). 

The vindex complex includes four species with either rugose or heavily 
punctured elytral interstriae. The rugosity of the elytra is strongest in vindex 
MacL. (Fig. 284) and difformis Lee; in igneus MacL. (= floridanus Ols.) 
and triangularis Say ( = torrens Lee), however, the elytral rugosity tends to 
be effaced and evident only with magnification as dense, confluencent punc- 
turing. All four species are predominantly coprophagous and commonly 
collected in cattle feces in pastures. P. igneus and triangularis are endemic to 
the United States (particularly the southeast) ; the ranges of vindex and 
difformis, while centered in the United States, extend into northern Mexico. 

The mexicanus complex brings together a relatively large number of 
species whose combined ranges extend from northwestern South America 
to the southwestern United States (Map 2). Most are endemic or largely 
endemic to tropical Mexico and adjacent Central America. All included 
species whose feeding habits are known are normally coprophagous al- 
though their ecological distribution (horizontal and vertical) appears quite 
wide. The mexicanus complex remains in need of revision at the species 
level using more characters than those employed here; a detailed study of 
these species including geographic variation and reference to type material 



Evolution of the Phanaeine Dung Beetles 833 

could well result in reduction to subspecific rank or synonymy of some of 
the species cited below. I have little doubt that the mexicanns complex is a 
natural group in spite of the fact that my study so far has produced no sound 
basis for assessing the interrelationships among the constituent clusters of 
obviously closely related species. I find that character states that are normally 
fixed in other phanaeine taxa of comparable size and/ or rank (e.g., various 
aspects of integumental sculpturing) vary in a seemingly irregular pattern 
among these clusters. Still the only satisfactory way, albeit taxonomically 
unsatisfying to me, of distinguishing closely related species is by differences 
in secondary sexual features of well developed males (particularly pronotal 
shape). The very striking phenetic similarity of small males and females of 
closely related species attests to a degree of virtual siblingness unapproached 
in any other phanaeine taxon. Identification of potentially sibling indi- 
viduals can at best only be inferred, and then only if they form parts of a 
sample including well developed males. 

Following are diagnoses of the species-clusters comprising the mexicanns 
complex; included are two isolated species treated as of cluster rank. In 
general, the species of each cluster resemble each other in at least the follow- 
ing respects: shape of male pronotum, relief of elytral interstriae (flat or 
convex); pronotal sculpturing (granulate or rugose) and great similarity of 
females. I have avoided compiling a key to species-clusters since variation 
of characters used would render it largely inefficient and little better than 
that of OlsoufiefT (1924: 37-46). 

Cluster 1 (lugens Nev., prasinus Har., hermes Har.) : Prothorax evenly 
granulate; male pronotum as in Figures 223-224; female pronotum convex, 
with three flat antero-median tubercles; elytral interstriae weakly convex, 
striae distinctly punctate (seen under 10X magnification). Besides striking 
interspecific female similarity, the males very closely resemble each other; 
the most conspicuous variation distinguishing most individuals of either sex 
is color. Both prasinus and lugens occur in Venezuela and adjacent regions 
of Colombia; hermes ranges from Colombia into Central America. 

Cluster 2 (mexicanns Har., amythaon Har., wagneri Har., lunaris 
Faschb.) : Prothorax rugose, less heavily so in female; male pronotum as in 
or similar to Figure 226; female pronotum as in Figure 227, with low-set, 
transverse antero-median ridge flanked by shallow, usually smooth con- 
cavities; elytral interstriae convex, striae simple or very weakly punctate. I 
cannot reliably identify females or small males of these species unless they 
are collected along with well-developed males of a single species; well- 
developed males differ in subtle features of pronotal shape. The distribution 
of this cluster is unique, the composite range extending from Peru to Ari- 
zona: lunaris occurs in eastern Peru (and Ecuador ?), wagneri is largely 



834 The University Science Bulletin 

Central American, mexicanus is very common in southern Mexico while 
amythaon has been collected from southwestern Mexico northward along the 
western seaboard of Mexico into Arizona (Prescott). 

Phanaeus achilles Bohart: I have seen one well developed male of this 
species; it resembles the males of cluster 2 in all respects except color (bright 
red above) and strong puncturing of the elytral striae. It is known from 
Peru and Ecuador. 

Cluster 3 {demon Har., obliquans Bates) : Pronotum rugose except dor- 
sally; male pronotum (see below); female pronotum as in Figure 227; elytral 
interstriae weakly convex, striae simple; leading margin of head distinctly 
notched at lateral extremity of each postclypeal carina (Fig. 109). The 
pronotum of the male of demon is unique in the genus (Fig. 225); the 
postero-lateral angles are produced as massive projections (which are acute 
and slanted posteriorly in obliquans) flanking a deep, smooth median con- 
cavity. Both species are ordinarily of a green color much more highly shining 
than the usually dull metallic colors of other species in the complex. P. 
demon is very common in the hot tropical valleys (e.g., of Rio Balsas) of 
southern Mexico; I know obliquans only from a few Central American 
localities. 

Cluster 4 (qimdridens Say, damocles Har.): Pronotum rugose; prono- 
tum of female convex, with transverse anteromedian ridge; elytral interstriae 
flat, shining, striae nearly effaced (Fig. 286). The pronotum of male 
quadridens is flat above with large, flaring postero-lateral angles; that of 
male damocles closely resembles that of cluster 5 (see below). Both species 
are evidently restricted to higher altitudes in Mexico; quadridens is character- 
istic of the Mexican Plateau (see HalfTter, 1964), where few other Phanaeus 
are encountered, while damocles is found in pine-oak forests in southern 
Mexico (Halffter, pers.comm.). 

Cluster 5 (palliatus Sturm, eximius Bates) : Pronotum rugose; male 
pronotum shaped as in Figures 223-224 except that dorsal surface concave 
with median tubercle; pronotum of female as in cluster 4; elytral interstriae 
convex, striae simple. This species pair is perhaps to be regarded as an 
annectent among clusters 5, 7 and 8. I have seen very few eximius, all from 
Central America; palliatus is also generally regarded as rare although it can 
be found in numbers in isolated localities in the mountains near Mexico City. 
G. Halffter and I collected many specimens of palliatus from burro and 
human dung near Ocoyoacac (state of Mexico); HallTter feels that it is 
narrowly distributed altitudinally. The variation in color of palliatus is 
striking; in the single Ocoyoacac sample are bright metallic red, green and 
dark blue individuals of both sexes. 

Cluster 6 (pilatei Har., sallei Har., beltianus Bates, flohri Nev.) : Pro- 
notum rugose; shape of pronotum of well developed males similar to cluster 



Evolution of the Phanaeine Dung Beetles 835 

1 but with very smooth, triangular area basally; pronotum of female convex, 
with small transverse antero-median ridge; elytral interstriae convex, striae 
simple. This cluster is evidently restricted to Central America and adjacent 
regions of Mexico. Although I have seen only a few specimens of each, I 
suspect that of the four names listed, only two will remain full species upon 
serious study. I know nothing of the biology of these species. 

Cluster 7 (guatemalensis Har., amethystinus Har., [probably = martinezi 
HalfTter|, scutijer Bates, melampus Har.): Pronotum rugose; pronotum of 
well developed males similar to Figures 223-224, with smooth basal area 
bearing small acute tubercle anteriorly; pronotum of female convex, with 
medially dentate ridge anteriorly; elytral interstriae convex, striae simple. 
These species are very similar; all except scutijer (likely a color variation of 
guatemalensis} are dull or blue-black. As in cluster 6, these species are re- 
stricted to southern Mexico and Central America. 

Phanaeus adonis Har.: This species, relatively common in valleys along 
the southern periphery of the Mexican Plateau (Halftter, pers. comm.), re- 
sembles triangularis more than any Mexican species. As in triangularis, the 
male pronotum is flat dorsally and strongly produced postero-laterally and 
the pronotum of both sexes is coarsely granulate. Unlike the vindex com- 
plex, however, the elytral interstriae are at most weakly punctured. 

Cluster 8 (daphnis Har., tricornis Ols., widens Cast., nimrod Har., 
jurcosus Bates): Pronotum rugose; pronotum of well-developed males with 
rounded postero-lateral angles, concave dorsally with usually large median 
or basal process of varying shape, sides of concavity serrate and/or dentate 
(Fig. 228); pronotum of female convex, with small antero-median ridge; 
elytral interstriae convex, striae simple. All these species are relatively small 
(to 15mm in length) and usually wholly green or yellow-green. I cannot 
reliably separate females and most small males of this cluster without cor- 
roborating evidence; of the subtle pronotal features distinguishing well 
developed males, the shape of the median process is most conspicuous. This 
cluster occurs from southern Mexico northward along the western coast of 
Sinaloa. 

Unassigned names probably belonging in the Phanaeus mexicanus com- 
plex: goidanichi Balths., pseudojurcosus Balths., substriolatus Balths., 
jran\enbergeri Balths., babori Balths. 

Genus Oxysternon Castlenau 

description 

Dorsal surface of clypeus completely transversely ridged or punctate 

posteriorly; (1) anterior margin of clypeus not emarginate medially, at most 

weakly toothed (Figs. 210, 213-214); (2) clypeal process transverse, spini- 

form (Fig. 112) or reduced; lateral clypeal carinae present or absent; (3) 



836 The University Science Bulletin 

cephalic process of male corniform (Figs. 210-211) or reduced to a short 
bituberculate process or two simple tubercles; (4) cephalic process of female 
a simple or tuberculate carina; (5) paraocular areas of parietals narrow; (6) 
shape of postocular lobes of parietals as described in part III, anterior portion 
of pronotal ridge effaced ( Mioxy sternon) or weakened or broken by shallow 
emarginations; canthal area distinct or not; (7) cephalic brachium bipodal 
(not observed for Mioxy stern on) \ (8) eyes small; (9) occipital ridge com- 
plete; (10) labrum Phanaeus-type; (11) mentum strongly emarginate; (12) 
median lobe of hypopharynx present; (13) premental sclerites completely 
sclerotized; (14) pronotum of male as in Figures 210-211 or 214-215; (15) 
pronotum of female type D (Figs. 212-213); (16) pronotum punctate (Fig. 
160), often appearing completely smooth; postero-median angle of pronotum 
distinctly produced as acute process fitting between inner basal angles of 
elytra (Fig. 210), posterior pronotal fossae absent (except ? biimpressum 
Ols.) ; antero-median angle of metastemum strongly produced as acute spine 
fitting between front coxae (Figs. 244-245) ; (17) front tarsi present in female; 
(18) front tibiae tridentate or quadridentate; (19) middle and hind tarsi five- 
segmented; (20) elytral interstriae smooth to punctate and (21) evenly 
convex (Figs. 280-281); (22) striae simple, sometimes nearly effaced; hind 
wing not notched; seventh abdominal spiracle marginal or intrasternal. 

In addition to the above features there is a tendency among species of 
Oxysternon for abdominal sternum 7 to be almost vertical relative to sternum 
6, such that, seen in profile, the abdomen has a decidedly angulate appearance. 

Key to the Subgenera and Species Groups of Oxysternon 

1. Cephalic process of male two closely-set median tubercles (Fig. 80), 

that of female either absent or a weak, sometimes tuberculate 
carina; clypeal process reduced to small tubercle; lateral clypeal 
carinae absent; occipital ridge nearly effaced; metasternum, along 
inner margin of each mesocoxa, with a single row of irregularly 
placed large punctures; elytral striae distinct, usually strongly 
punctate; last abdominal spiracle intrasternal; color metallic blue 

or green; length 12-15mm Mioxysternon 

-Cephalic process of male horn-like, often massive or, in small indi- 
viduals, reduced to a pair of closely set median tubercles; that of 
female a strong, trituberculate carina; clypeal process transverse 
or spiniform; lateral clypeal carinae present, at least partially; 
occipital ridge strongly developed; metasternum not strongly punc- 
tate adjacent to mesocoxae; elytral striae present but often nearly 
effaced, never strongly punctate; last abdominal spiracle marginal; 
color variable, usually bright; length 10-35mm Oxysternon , sen. str 2 

2. Dorsal surface of clypeus black; clypeal process spiniform; vertical, 

postero-lateral portion of circumnotal ridge effaced by a broad, 
oblique thickening (Fig. 211, arrow); each mesocoxa paralleled 
along outer side by wide, raised ridge-like portion of metasternum 



Evolution of the Phanaeine Dung Beetles 



837 




Map 3. Approximate distributions of the subgenera of Oxysternon. (?'s refer to occurrence 
of Oxysternon, sen. str.) 



which does not bend dorsally at anterior end; front tibiae tridentate, 
only apical two teeth carinate on anterior surface; hind tibiae each 
with weak transverse carina on outer face near apex; hind apical 

tibial spur truncate distally jestivum group 

Dorsal surface of clypeus at least partially metallic; clypeal process 
transverse (spiniform in palaemon); circumnotal ridge not effaced 
postero-laterally; thickened outer border of mesocoxal depression 
narrow, bent dorsally at anterior end; front tibiae quadridentate, 
apical three teeth carinate anteriorly, basal tooth small; hind tibiae 
not carinate on outer apical surface; hind tibial spur acute .... silenum group 



838 The University Science Bulletin 

Subgenus Oxysternon Castlenau, sensu stricto 

Oxysternon Castlenau, 18-10: 82. Type: Scarabaeus festivus L., present designation. 

Sternaspis Hope, 1837: 52 (junior homonym of Sternaspis Otto. Type: Scarabaeus festivus L., 

present designation. 
Strombodes Gistel, 1857: 602. Type: Scarabaeus festivus L., original designation. 

Among the members of this subgenus are the better known Oxysternon. 
All are large, usually conspicuously colored South American phanaeines 
(Map 3). Both constituent species groups are quite cohesive, although 
palaemon is a phenetic annectant between them (see below). 

The festivum group (festivum |L.|; macleayi Nev.; lautum |MacL.]; 
conspicillatum [Web.]; ebeninum Nev.): Various structural details (par- 
ticularly the shape of the pronota of well-developed males) can be used to 
subgroup the constituents of this taxon. Thus, festivum and macleayi form 
a species pair, and the remaining three species constitute a second subgroup. 
All species have characteristic coloration, which evidently varies little. Two, 
festivum and conspicillatum, are quite common in collections. Surprisingly 
little is known of the biology of this group, although collection data indicate 
all are amazonian. The range of conspicillatum is enormous, covering the 
entire Amazon Basin south of the Amazon and Rio Negro and certain non- 
basin areas (e.g., Quito, Ecuador). 0. festivum is also amazonian but occurs 
north of the Amazon River (? into the Guianas); festivum has been collected 
from traps baited with human feces. The known range of ebeninum is 
restricted along the upper Rio Negro. Both lautum and macleayi are known 
only from the Amazon Basin. 

The silenum croup (silenum Cast.; smaragdinum Ols.; sericeum Ols.; 
palaemon Cast.): The most common species, palaemon, is noteworthy in 
that it shows features (e.g., spiniform clypeal process) otherwise unique to 
the festivum group. This species is coprophagous and occurs in great num- 
bers in the savanna regions ("cerrados") of south-central Brazil. I have seen 
very few specimens of silenum, smaragdinum and sericeum; they appear so 
similar to me as to suggest conspecificity. All specimens of these three species 
that I have seen were collected at scattered localities in the western portions 
of the Amazon Basin. 

Unassigned names probably Oxysternon, sen. str.: biimpressum Ols., 
aeneum Ols. 

Mioxysternon new subgenus 

Type: Oxysternon spinijerum Cast. (Derivation of subgeneric name: duos (dwarf) -f- Oxy- 
sternon, neuter; reference to relatively small size.) 

This subgenus brings together three small-sized, rare species of Oxy- 
sternon (spiniferum Cast.; curvispinum Ols.; pteroderum Nev.) clearly 
distinguished from their congeners by those features mentioned in the key. 



Evolution of the Phanaeine Dung Beetles 839 

Most noteworthy of these are the apparent lack of a corniform cephalic 
process on otherwise well developed males, small size, obliteration of the 
occipital ridge and shape of the pronotum of the male (Figs. 214-215), which 
in the type-species resembles that of Phanaeus bispinus. The pronotum of 
the single male pteroderum I have seen bears two parallel, longitudinal crests 
rather than the parallel prolongations of spiniferum. Of interest is the fact 
that the cephalic process of the male of Mioxystemon resembles that charac- 
teristic of the "minor" males of Oxysternon, sen. str., two closely-set tubercles. 
I have seen very few specimens of these species, all from scattered South 
American localities (Map 3). 

Judging by original description, striatopunctatum Ols. would also fall in 
this subgenus. 



J b^ 



Genus Coprophanaeus Olsoufiert 

DESCRIPTION 

(1) anterior margin of clypeus emarginate medially (Figs. 83-88); (2) 
clypeal process U-shaped (Fig. 114); lateral clypeal carinae present; (3) 
cephalic process of male corniform or lamellate (Figs. 181, 186, 188, 192-193) ; 
(4) cephalic process of female corniform (Fig. 182) or carinate; (5) para- 
ocular areas of parietals usually narrow (Fig. 116); (6) postocular lobes of 
parietal depressed transversely, pronotal margin entire (Fig. 116) or weakly 
effaced but never distinctly broken by emarginations; canthal area usually 
distinct (Fig. 116A); (7) cephalic brachium monopodal (Fig. 126); (8) 
eyes almost always large; (9) occipital ridge incomplete; (10) labritm 
Coprophanaeus-/y/?<? (Figs. 146-148); (11) mentum weakly emarginate (Fig. 
157); (12) median lobe of hypopharynx present; (13) premental sclerites 
marginally sclerotized (as in Fig. 150); (14) pronotum of male type C, E or 
F (Figs. 181, 183, 186, 189, 192); (15) pronotum of female type A (Figs. 
184-185, 190-191) or like type E of male (Figs. 181-182); (16) pronotum 
usually at least partially rugose or distinctly ridged (Fig. 166) ; posterior 
pronotal fossae usually present; (17) front tarsi present or absent in female; 
(18) front tibiae Coprophanaeus-/y/7<? (Fig. 246); (19) middle and hind tarsi 
five-segmented; (20) sculpturing of elytral interstriae variable, usually 
smooth; (21) interstriae flat or convex; (22) striae sometimes carinulate; 
hind wing sometimes notched basally (Fig. 292) ; position of last abdominal 
spiracle variable. 

This genus brings together three of the five subgenera of Phanaeus estab- 
lished by Olsoufiert (1924): Coprophanaeus, Megaphanaeus and Metallo- 
phanaeus. The subgeneric rank of these three taxa is here maintained. All 
Coprophanaeus are immediately recognizable by the presence of emarginate 
clypeus, strongly quadridentate front tibiae and usually distinct ridging of 



840 The University Science Bulletin 

the pronotum. The only other phanaeines likely to be confused with this 
genus are the few robust species of Dendropaemon. Coprophanaeus is South 
American although a few species occur in Mexico (Maps 4 and 5) ; all species 
whose habits have been observed are exclusively or ordinarily necrophagous, 
usually nocturnal and tend to occur in forests. 

Key to the Subgenera and Species Groups of Coprophanaeus 

1. Front tarsi present in female; striae conspicuous, always distinctly 

carinulate, carinulae straight or undulate (Fig. 288); hind wing 
not notched basally; last abdominal spiracle located within sternum 
7 (though not visible laterally in bellicosus); posterior surface of 
each front tibial tooth with oblique, conspicuous brush-like clump 
of short setae; female cephalic process (except of bellicosus) and 
pronotal shape very similar to those of male such that sexual 
dimorphism, especially among smaller individuals, is not pro- 
nounced; pronotum of both sexes with an often massive, more-or- 
less saddle-shaped or quadrate process extending anteriorly from 
posterior portion of large median concavity; paraocular areas of 
parietals dorsally distinctly carinate lateral to eye; occipital areas of 
parietals each with angulate prominence; size large to very large 
(40-60mm); dorsum dark metallic blue, green or deep violet 
Megaphanaeus 

-Front tarsi absent in female; striae often nearly effaced, at most only 
weakly carinulate in some individuals with green elytra; hind 
wing notched or not basally; last abdominal spiracle marginal or 
set in distinct notch in edge of sternum 7; posterior surface of front 
tibial teeth each with single, oblique row of setae; sexual di- 
morphism, as expressed by cephalic and pronotal processes, very 
pronounced or not; pronotum of female with shallow antero- 
median concavity bordered anteriorly by transverse, crest-like 
ridge; pronotum of male variable; paraocular areas of parietals 
with no distinct carina lateral to eye (except in pertyi and pessoai); 
occipital areas of parietals more-or-less flattened, without distinct 
angulate prominence; size moderate (20-40mm); color variable 2 

2. Hind wing not notched basally; width of lower portion of each eye less 

than twice (usually about equal) width of adjacent oculogular 
space; cephalic process of male a simple, tapering "horn"; color of 
pronotum and usually also elytra metallic, shining Metallophanaeus 

- Hind wing distinctly notched basally, vein V 2 usually conspicuously 
hooked apically (Fig. 292); lower portion of eyes large, bulging, 
width of each more than twice width of adjacent oculogular 
space; cephalic process of male lamellate, as in Figures 192, 193 or 
188, never a simple, tapering horn; dorsum usually black or dull 
blue-black with metallic areas restricted to margins of pronotum, 
etc., rarely nearly wholly metallic green above (C. spitzi and some 
specimens of C. milon and pluto) Coprophanaeus, sen. str 3 

3. Anterior portion of circumnotal ridge not distinctly effaced behind 



Evolution of the Phanaeine Dung Beetles 841 

eyes; last abdominal spiracle marginal, not set in distinct notch in 
sternum; posterior margin of paraocular areas of parietals (seen 
dorsally) straight, not bending toward posterior angle of eye; 
pronotum of large males never with strongly produced, quadrate 
median process; cephalic process of larger males as in Figures 192- 
193, its sides distinctly convergent apically jasius group 

- Anterior portion of circumnotal ridge distinctly effaced behind eyes, 
gaps so produced each as wide or wider than adjacent portion of eye 
(as in Fig. 119C); last abdominal spiracle set in distinct notch in 
dorsal margin of sternum 7; posterior margin of paraocular areas 
seen dorsally) distinctly bent toward posterior angle of eye; pronotum 
of well-developed males with large quadrate median process (Figs. 
186-187); cephalic process of male with sides parallel or divergent, 
usually plate-like with two or three apical tuberculate processes 
(Fig. 188) dardanus group 

Subgenus Megaphanaeus OlsoufiefT 

Megaphanaeus Olsoufieff, 192-1: 23 (as a subgenus of Phanaeus). Type: Scarabaeus lancijer 
L., original designation. 

This subgenus includes some of the largest known Scarabaeinae; the 
largest phanaeines, C. ensijer and lancijer, regularly exceed 50mm in length; 
only some species of Heliocopris are larger. Also noteworthy is the strong 
resemblance of males and females in development of secondary sexual fea- 
tures, the most unusual aspect of which is the presence of a massive cephalic 
"horn" on the female. The sexual resemblance is only superficial since sepa- 
ration of sexes can always be immediately accomplished by examination for 
front tarsi, which are present only in females. 

The name Megaphanaeus is used here as established by Olsoufieff. It 
includes four species {lancijer [L.|; ensijer Germ.; bonariensis Gory; belli- 
cosus Oliv.), the first three of which are evidently rather closely related. 
These three species have strongly sculptured elytra with undulate strial 
carinulae; the pronotum is likewise strongly sculptured, bearing flattened 
granules which coalesce to form extensive transverse ridging. C. bellicosus 
differs from them in several respects: non-corniform cephalic process of the 
female (which closely resembles that of Sulco phanaeus carnijex, Figs. 201- 
202), smooth, flat elytral interstriae, straight elytral carinulae and deep violet 
color. (The above species are bright metallic blue, green or blue-green.) 
Among other features, elytral sculpturing can be used to separate the species 
of this subgenus. 

With the possible exception of bellicosus, whose habits I do not know, 
Megaphanaeus is a necrophagous group. I have collected lancijer in sur- 
prisingly large numbers from carrion-baited traps along the Amazon River 
(Belem and Manaus) and ensijer in a similar manner in the state of Sao 
Paulo, Brazil. Martinez (1959) reports bonariensis to be necrophagous, 



842 



The University Science Bulletin 



& 




MAP 4 



> Metal lophanaeus 



: : : :l]: : : : . Megaphanaeus 



Map 4. Approximate distributions of the subgenera Megaphanaeus and Metallophanaeus. 

preferring bird carcasses. All species are South American (Map 4), where 
they appear to be restricted to forests. 



Subgenus Metallophanaeus OlsounefT 

Metallophanaeus Olsourteff, 1024: 22 (as subgenus of Phanaeus). Type: Phanaeus saphirinus 
Sturm, original designation. 

This subgenus brings together two species-pairs, saphirinus Sturm- 
horus (Waterhouse) and pertyi Oh.-pessoai Pereira. According to Martinez 
and Pereira (1967), thalassinus Perty, which I have not seen, is closely re- 



Evolution of the Phanaeine Dung Beetles 843 

lated to the latter pair. Both pairs are quite distinct and perhaps brought 
together only artificially here, the only alternative (which I consider inade- 
quate) being to erect a separate subgenus for pertyi-pessoai. Olsoufieft (1924) 
and Martinez and Pereira (1967) include the latter pair in Coprophanaeus, 
sen. str. The general appearance of the pertyi group suggests an annectant 
taxon between Metallophanaeus and Coprophanaeus, sen. str. All species 
except pertyi occur in southern South America (Map 4) ; pertyi is limited to 
the arid regions ("caatinga") of northeast Brazil. Contrary to Halftter and 
Matthews (1966:30) saphirinus is at least sometimes necrophagous (Ed- 
monds, 1967) ; the same authors cite horns as necrophagous. 

Subgenus Coprophanaeus Olsoufieft", sensu stricto 

Coprophaneaus Olsoufieft, 1924: 22 (as subgenus of Phanaeus). Type: Scarabaeus jasius Oliver, 
original designation. 

The name Coprophanaeus is used here in a somewhat more restricted 
taxonomic sense than by Olsoufieft. The included species, which comprise 
the bulk of the genus, are all somber colored and robust and at a glance 
rather closely resemble each other. 

The jasius group includes at least four species (jasius [Oliver]; cerberus 
|Har.]; parvulus [Ols.]; spitzi [Pessoa]) all restricted to South America 
(Map 5) ; jasius occupies an enormous range coincident with all but the 
southernmost regions of the composite range of the group. I have collected 
all four included species from carrion-baited traps in Brazilian forests. 

The dardanus group also comprises at least four species (milon [Blanch.]; 
telamon [Er.]; dardanus [MacL.]; pluto [Har. |). The composite range of 
this group is much wider than the former. This is the only coprophanaeine 
taxon species of which occur outside South America; i.e., the Central Amer- 
ican distribution indicated in Map 5. 

Unassigned names probably belonging in Coprophanaeus, sen. str.: cos- 
tatus Ols., arrowi Ols., jasion Fels., boucardi Nev., punctatus Ols., cono- 
cephalus Ols., ohausi Fels., sericeus Fels., rex Balths., strandi Balths., roubali 
Balths. 

Genus Dendropaemon Perty 

DESCRIPTION 

(1) anterior margin of clypeus emarginate medially (Figs. 89-94, 132, 
134) ; (2) clypeal process U-shaped, sometimes reduced; lateral clypeal 
carinate present or absent; (3) cephalic process of male usually carinate, 
sometimes lamellate (Fig. 305); (4) cephalic process of female carinate; 
paraocular areas of parietal s usually raised medially into large bump or 
transverse ridge (Fig. 134) ; (5) paraocular areas of parietals usually wide 
(Fig. 134) ; (6) postocular lobes of parietals depressed transversely, circum- 



844 



The University Science Bulletin 




i/&K- 



'tr 




^> 



■%> . 



MAP 5 



Coprophanaeus, sen. str. 




&F 



Map 5. Approximate distribution of Coprophanaeus, sen. str. 



notal ridge not interrupted by emarginations (Figs. 123C, 124C) ; canthal 
area distinct (Figs. 123 A, B, 124A, B); (7) cephalic brachium monopodal; 
(8) eyes small or large; (9) occipital ridge incomplete; (10) labnim Copro- 
phanaeus-ty/?^ (Figs. 144-145), apical margin deeply emarginate, median 
brush reduced to jew long setae inserting in apical for\ of antero-median 
process of suspensorium; (11) mentum wealth emarginate (Fig. 152); (12) 



Evolution of the Phanaeine Dung Beetles 845 

median lobe of hypopharynx absent; (13) premental sclerites marginally 
sclerotized (Fig. 151); (14-15) pronotum of both sexes usually similar, more- 
or-less flattened (see discussion below); (16) pronotum partially roughened 
or entirely punctate; posterior pronotal fossae present; transverse propleural 
carina present {Dendropaemon, sen. str., Figs. 179-180) or absent (Copro- 
phanaeoides and Paradendropaemon); (17) front tarsi absent in female; 
(18) front tibiae Dendropaemon-/y/7(? (Fig. 247) (except ganglbaueri); 
lower apical angle of middle and hind femora expanded, covering base of 
tibia (Figs. 268-269) (except ganglbaueri); (19) middle and hind tarsi two- 
or three-segmented (Figs. 270-279) ; (20) elytral interstriae smooth or punc- 
tate; (21) same evenly convex; (22) striae carinulate (Figs. 290-291) (very 
weakly so in ganglbaueri); hind wing notched (Figs. 294-295, 298) or not 
(Fig. 299) ; base of pygidium impressed on each side of median furrow (Fig. 
162, Id) (very weakly so in ganglbaueri). 

In addition to the above features, Dendropaemon is characterized by a 
remarkably great variation in body shape. Included in the genus are two 
species (amyntas and ganglbaueri) of the generally robust shape character- 
istic of other phanaeines (Figs. 304-305). Most Dendropaemon, however, 
are narrowed and dorso-ventrally flattened to some degree (Figs. 306-309). 
Following is a list of some of those observed skeletal features judged directly 
related to the flattening-narrowing trend among these species. As indicated, 
most vary continuously among progressively flatter, narrower species. 

a. Reduction in size of cephalic process of male; prominent in robust 
species (Fig. 305), carinate in flattened species (Figs. 307, 309). 

b. Dorso-ventral flattening of antennal club (Fig. 136A) in all species 
except ganglbaueri; in some (Fig. 136) flattening accompanied by elongation 
of first lamella (segment 7). 

c. Maximal elongation of mentum, approximately as long as wide in all 
species (Fig. 152). 

d. Progressive dorso-ventral flattening of prothorax, from robust (Figs. 
167-168, 174) to successively more flattened (Figs. 170, 173, 175-176); flatten- 
ing is accompanied by progressive narrowing and squaring of dorsal outline, 
from rounded laterally (Figs. 169, 304) to more parallel-sided (Figs. 171, 177, 
306, 308) and by reduction of pronotal processes (cf., Figs. 167, 176). 

e. Narrowing of mesoscutellum; in all Dendropaemon length at least 
equal to, but usually greater than width (Fig. 229); in other phanaeines, 
width always greater (usually much greater) than length (Fig. 44). 

f. Flattening and elongation of metepisternum; in other phanaeines and 
in robust Dendropaemon, length (along dorsal margin) less than 1.5 times 
height (along anterior margin) (Figs. 54, 241, 244); in flattened species, 



846 



The University Science Bulletin 




306 



308 






Figs. 304-309. 304, Dendropaemon amyntas, male, dorsal view; 305, same, lateral view; 
306, D. viridis, male, dorsal view; 307, same, lateral view; 308, D. nitidicollis, male, dorsal 
view; 309, same, lateral view. 



length progressively proportionately greater, but always more than twice 
height (Figs. 232, 235, 238, 242-243). 

g. Narrowing of metasternum; in robust species, length of its median 
area (from antero-median angle to posterior margin) only slightly greater 
(less than 1.25 times) than width (smallest distance between mesal margins 



Evolution of the Phanaeine Dung Beetles 847 

of coxal depressions) (Fig. 51); in narrowed species, length progressively 
proportionately greater but always twice or more width (Figs. 231, 234, 237); 
accompanying narrowing is progressive reduction in prominence of antero- 
median angle of metasternum (Figs. 230, 233, 236). 

h. Narrowing of elytra; in robust species, combined widths of elytra (at 
humeral angles) greater than length (along elytral suture) (Fig. 304); in 
narrowed species, length greater than width (Figs. 306, 308); narrowing 
accompanied by proportionate narrowing of abdomen, which is covered by 
elytra. 

Primarily because of the great variation in general form it is difficult to 
recognize Dendropaemon immediately using superficial "sight" criteria. 
Most species are more-or-less flattened, elongate and parallel-sided; all have 
a reduced number of tarsomeres (two or three). The extreme of the elon- 
gation-flattening trend observable in smaller species of the genus is paralleled 
in Onthocharis Westwood which superficially resembles small Dendro- 
paemon also in color (metallic green) and clypeal emargination, and in 
Stiptopodiits Har. 

A meaningful evolutionary interpretation of the remarkable variation in 
form of Dendropaemon is largely precluded by a lack of reliable data on its 
biology, which likely will prove to be quite aberrant. The genus name 
means "tree dweller," an appellation of doubtful accuracy. OlsoufiefT (1924: 
121) and Blut (1939:263) state emphatically that Dendropaemon inhabit 
rotting wood; Halffter and Matthews (1966:40) suggest they may feed on 
fungi in rotting wood. Blut (p. 266) goes so far as to say that given species 
of Dendropaemon feed selectively in given species of hardwood trees! 
Evidently, however, the wood-dwelling habits of the genus have yet to be 
positively demonstrated. All specimens about which I could learn circum- 
stances of capture (from collector or label) were collected at lights or lying 
dead on the ground. Moreover, neither I nor any other collector whom I 
know personally has been able to collect adults from rotting logs, etc., in 
localities where they are known to occur. To my knowledge, Dendropaemon 
has likewise never been collected along with other phanaeines from carrion 
or feces of any kind. That some (or all) species could be dendrophilous is, 
of course, possible; but other, equally possible habitats have not yet been 
adequately explored, such as vertebrate burrows and nests, nests of ground- 
dwelling termites and ants, and fungi (particularly hymenomycetes). 

Most museum specimens of Dendropaemon are quite old and, hence, apt 
to bear non-precise locality data. Consequently, Map 6 is at best a tenuous 
approximation of the range of the genus. The detailed distributional com- 
ments of Blut (1939:265) need confirmation. 

A detailed revision of Dendropaemon species is greatly needed. The 



848 



The University Science Bulletin 




A Megatharsis 
• Tetramerei 

x'f.y.: Dendropaem 



Map 6. Approximate distributions of Megatharsis, Tetramereia and Dendropaemon. (?'s 
refer to occurrence of Dendropaemon.) 



descriptions and keys of Olsoufieff (1924), Pessoa and Lane (1936) and Blut 
(1939), which in general were based on too few specimens, need enlarge- 
ment and clarification. Blut lists twenty-five species, fifteen of which I have 
examined in addition to two others of uncertain name. Heretofore the genus 
has been informally divided into two species groups based on number of 
tarsal segments (two or three). In the following treatment tarsomere num- 
ber is de-emphasized since other, more reliable variation quite clearly 
indicates an altogether different systematic picture. 



Evolution of the Phanaeine Dung Beetles 849 

Key to the Subgenera and Species Groups of Dendropaemon 

1. Abdominal sternum 7 long, as long as combined lengths of sterna 5 

and 6 (Fig. 300); inner basal angle of front tibiae not expanded 
as in Figure 247, tibiae gradually narrowing basally; hind tarsi two- 
segmented, basal segment cylindrical, nearly as long as tibia, apical 
segment minute, hemispherical (Fig. 278, arrow indicates apical 

segment); color dull black-brown Paradendropaemon 

-Abdominal sternum 7 no longer (usually shorter) than sternum 6 
alone; front tibiae as in Figure 247; hind tarsi two- or three-seg- 
mented, basal segment flattened, its length never approaching that 
of tibia; color variable 2 

2. Transverse propleural carina absent; eyes large (Figs. 132-133); tarsi 

three-segmented (Fig. 274) Coprophanaeoides 

-Transverse propleural carina present (Figs. 179-180, arrows); eyes 
small (Fig. 134; large in denticollis); tarsi two- or three-segmented 
Dendropaemon, sen. str 3 

3. Transverse propleural carina complete, extending laterally beyond coxa 

(Fig. 180); emarginations of anterior clypeal margin variable (Figs. 
90, 92-93); tarsi three-segmented (Figs. 273, 275-276); hind wing 

not notched (Fig. 299); length 8-15mm viridipennis group 

-Transverse propleural carina incomplete, not extending laterally from 
covn (Fig. 179); emarginations of clvpeal margin as in Figures 94 
and 134; tarsi two- or three-segmented; hind wing notched or not 4 

4. Middle and lateral emarginations of clypeus rounded, teeth rounded 

(Fig. 94); tarsi three-segmented (Figs. 270-271; two-segmented in 
smaragdinum , Fig. 272); hind wing at most only slightly notched 
(Fig. 298); length 6-9mm; flattened, elongate, metallic green 
species denticollis group 

- Middle emargination of clypeus acute, lateral emarginations absent, 
teeth acute (Fig. 134); tarsi two-segmented (Fig. 277), length of 
basal segment more than three times length of apical segment; hind 
wing deeply notched (Fig. 295); length 7-12mm; flattened, solid 
black or metallic green species plceum group 

Paradendropaemon new subgenus 

Type: Dendropaemon ganglbaueri Felsche. (Derivation of subgeneric name: para [beside] + 
Dendropaemon, neuter; reference to taxonomic isolation within genus.) 

I have seen only one specimen of Dendropaemon ganglbaueri (a female), 
but it is so unlike its congeners that I must conclude that it warrants place- 
ment in a distinct subgenus. The following are those features which con- 
tribute to its uniqueness: (1) very large, triangular clypeal teeth (Fig. 91), 
(2) evenly tapering front tibiae (not strongly expanded basally as in other 
Dendropaemon, Fig. 247), (3) middle and hind femora not expanded 
apically as in other Dendropaemon (Figs. 268-269), (4) convex, evenly 
granulo-rugose pronotum (very similar to that of Tetramereia and Homalo- 
tarsus), (5) unique hind tarsi (Fig. 278), (6) exceptional length of ab- 



850 The University Science Bulletin 

dominal sternum 7 (Fig. 300), which is proportionately much greater than 
in any other phanaeine, and (7) only very weak, scarcely visible lateral de- 
pressions of base of the pygidium (distinctly strongly depressed in all other 
Dendropaemon). It is, of course, possible that any of the above features may 
be limited to the female. This robust species is known so far only from a 
few localities in the state of Sao Paulo, Brazil. 

Coprophanaeoides new subgenus 

Type: Dendropaemon renatii OlsoufiefF. (Derivation of subgeneric name: Coprophanaeus + 
-aides [resembling], neuter; reference to resemblance to Coprophanaeus, sen. sir.) 

This subgenus brings together at least three species, the names of only 
two of which I can use confidently: renatii Ols. and hirticollis Ols. I have 
studied a third species under the name of bahianum Har., considered by 
Blut (1939: 295-296, q.v.) as of incertae sedis because of the confusion attend- 
ing its original description and subsequent usage. These three species bear 
striking resemblance to Coprophanaeus, sen. str., particularly because of the 
shape of the medial emargi nation of the clypeus (Fig. 132), large eyes (Figs. 
132-133), ridged pronotal sculpturing and deeply emarginate hind wings. 
Unlike any other phanaeine, the entire upper surface of hirticollis is covered 
by scattered, erect setae. The distributional data that I have indicate that the 
range of Coprophanaeoides is centered in southern Brazil. 

Subgenus Dendropaemon Perty, sensu stricto 

Dendropaemon Perty, 1830: 30. Tvpe: Dendropaemon piceum Perty, subsequent designation by 

Blut, 1939: 267. 
Enicotarsus Castlenau, 1831: 35. Type: Enicotarstis riridipennis Cast., by monotypy. 
Onthoecus Dejean, 1837: 156. Type: Dendropaemon amyntas Har., present designation. 

The bulk of the species studied fall into this subgenus, the most salient 
feature of which is the presence of propleural carinae. The species included 
here express the entire range of body shape variation described above, the 
extremes of which are shown by D. amyntas (Figs. 304-305) and D. nitidi- 
collis (Figs. 308-309). 

The viridipennis group comprises four rather large (15-22mm) species 
bearing complete transverse propleural carinae and three-segmented tarsi 
(viridipennis [Cast.]; amyntas Har.; jractipes Felsche; lobatum Waterh.). 
D. amyntas, a robust species, superficially resembles "normal" phanaeines, 
and, besides ganglbaueri, it is the only Dendropaemon likely to be so con- 
fused. D. lobatum is a subcylindrical species which, although parallel-sided, 
presents cephalic and pronotal processes similar to those of amyntas. The 
other species, while large, are distinctly flattened and parallel-sided. D. 
viridipennis bears a striking resemblance to viridis in size, shape and color. 

The denticollis group here includes five, very similar small (7-llmm) 
metallic green species (denticollis Felsche, nitidicollis Ols., fasces Blut, 



Evolution of the Phanaeine Dung Beetles 851 

haroldi Ols., smaragdinum Waterhouse). All have distinctly rounded clypeal 
teeth and emarginations (Fig. 94) and incomplete propleural carinae. D. 
nitidicollis (Figs. 308-309) is the smallest known phanaeine (length 7mm). 

The piceum group comprises three species with incomplete propleural 
carinae, single emargination of clypeus with large, acute teeth (Figs. 90, 134) 
and two-segmented tarsi (Fig. 277). The hind wing of these species is dis- 
tinctly deeply notched (Fig. 295) (piceum Perty, telaphum Waterhouse, 
viridis Perty). 

Unassigned names probably Dendropaemon sen. str. (based on Blut, 
1939) : redugens Waterhouse, angustipennis, Har., silvanum Blut, batrachites 
Blut, similis Blut, subcylindricum Blut, temtitarsis Felsche, crenato striatum 
Felsche, quadratum Cast. 

Genus Tetramereia Klages 

Tetramereia Klages, 1907: 141. Type: Tetramereia jredcrichji Klages = Dendropaemon con- 

vextim Har., original designation. 
Enrypodea Klages, 1906: 1. Type: Eurypodea jrederickj Klages = Tetramereia jrcderickji 

Klages = Dendropaemon convexum Har., by monotypy. 
Boucomontins OlsoufiefT, 1924: 120. Type: Dendropaemon convexum Har., by monotypy. 

Tetramereia convexa (Har.) (Figs. 312-313) is a robust, brown species 
with an involved nomenclatorial history during which it has been placed in 
four genera (as type-species of three!). Klages (1906) proposed the name 
Eurypodea frederic\i (n. gen., n. sp.) in a single page article printed at his 
own expense. A series of errors resulted in the publication of another de- 
scription of the same species proposing the name Tetramereia fredericlui 
(Klages, 1907). Both descriptions were based on the same specimen. Olsou- 
fiefT (1924: 120), suspecting the duplication, nevertheless proposed still an- 
other genus name, Boucomontins , for Dendropaemon convexum Har., later 
found to be a senior synonym of T. frederic\ii. I have elected for now to 
continue usage of the name Tetramereia for two reasons: 1) I have judged 
Eurypodea does not comply with Articles 7 and 8 (Chapter IV, Criteria of 
Publication) of the International Code of Zoological Nomenclature and, 
hence, is unavailable (Art. 10) (see also Janssens, 1940) ; 2) Tetramereia is 
the name now of common usage. Nevertheless, I have requested that the 
International Commission on Zoological Nomenclature render an opinion 
on the availability of the name Eurypodea. 

DESCRIPTION 

(1) anterior margin of clypeus not emarginate medially, distinctly toothed 
(Fig. 95); (2) clypeal process reduced, a small, very weak transverse ridge; 
lateral clypeal carinae absent; (3-4) cephalic process of both sexes a simple 
transverse carina; paraocular areas of parietals flat and (5) wide; (6) post- 
ocular lobes of parietals depressed transversely, anterior portion of circum- 



852 The University Science Bulletin 

notal ridge not interrupted by emarginations; canthal area distinct; (7) 
cephalic brachium monopodal; (8) eyes small; (9) occipital ridge incom- 
plete; (10) labrum resembling Coprophanaeus-type (Figs. 142-143); (11) 
mentum weakly emarginate (Fig. 153); (12) median lobe of hypopharynx 
absent; (13) premental sclerites marginally sclerotized; (14-15) pronotum 
of both sexes evenly convex except for small, antero-median transverse carina; 
(16) pronotum evenly granulo-rugose; posterior pronotal fossae absent; 
mesometasternal suture angulate (as in Fig. 240); mesoscutellum as long as 
wide (as in Fig. 239); (17) front tarsi absent in female; (18) front tibiae 
quadridentate, apical three teeth carinate (Fig. 249); (19) middle and hind 
tarsi jour-segmented (Fig. 263) ; (20) elytral interstriae weakly punctured, 
(21) evenly convex; (22) striae simple; hind wing not notched (Fig. 297); 
last abdominal spiracle marginal; length 13-15 mm. 

So far this species is known only from a few scattered South American 
localities (Map 6) ; the type locality of frederic{ii is Suapure, Cuara (River) 
Valley, Venezuela. Tetramereia convexa very closely resembles Homalo- 
tarsus impressus (see discussion of Homalotarsus below). It is the only 
known phanaeine with four-segmented tarsi, a feature which indicated to 
Blut (1939) and Olsouneff (1924) an intermediate phylogenetic position of 
the genus between Dendropaemon and Megatharsis (see phylogenetic dis- 
cussion in the last part of this paper). Because of the lack of developed 
cephalic and pronotal protuberances, the sexes of this species very closely 
resemble each other. The biology of Tetramereia is unknown. 

Genus Homalotarsus Janssens 

Homalotarsus Janssens, 1932: 119. Type: Homalotarsus impressus Janssens (by original desig- 
nation and monotypy). 

To my knowledge only two specimens of this curious phanaeine are 
known, both of which are deposited in the Institut Royal des Sciences 
Naturelles de Belgique (Brussels). The holotype is labeled "South Amer- 
ica"; the second specimen, "Brazil." I was able to examine only the latter 
(sex unknown; Figs. 310-311); it bears Janssens' determination label and 
agrees with his original description of the genus. Homalotarsus impressus is 
remarkably similar to Tetramereia convexa, and the two should likely be 
considered congeners. Nevertheless, until more is known about both species, 
I choose to maintain the generic distinction. 

I was unable to examine the mouthparts and endoskeletal features of H. 
impressus; I have assumed that they are similar to their counterparts in 
Tetramereia. Moreover, because of their great similarity, I have restricted 
my description of Homalotarsus to the following observed features which 
differ markedly from their counterparts in Tetramereia; those features not 
mentioned are as in T. convexa (see above) : 



Evolution of the Phanaeine Dung Beetles 



853 





312 



313 






315 




Figs. 310-315. 310, Homalotarstis impressus, female?, dorsal view; 311, same, lateral view; 
312, Tetramereia convexa, female, dorsal view; 313, same, lateral view; 314, Megatharsis 
buckjeyi, female, dorsal view; 315, same, lateral view. 



854 The University Science Bulletin 

1. shape of mentum and labial palpi as in Figure 156; 

2. shape of front tibia as in Figure 254, apical tooth not carinate; 

3. middle and hind tarsi five-segmented (Figs. 264-265); 

4. dorsal surface of apex of hind femur, of tibia and of first four tarso- 
meres bearing large setigerous granules (Fig. 260); 

5. color metallic yellow-green. 

Genus Megatharsis Waterhouse 

Megatharsis Waterhouse, 1891: 59. Type: Megatharsis buckleyi Waterhouse, by monotypy. 

Although Waterhouse (1891) likened this species to Bolbites onitoides 
Har., Megatharsis buc/(leyi (Figs. 314-315) is a unique phanaeine which, to 
my eye, resembles Bolbites very little if at all. I have examined only two 
specimens (females), both collected in Macas, Ecuador; the type locality is 
nearby Chiquinda, Ecuador (Map 6). The following description is based on 
females although I fully expect the male to agree closely with it; Waterhouse 
failed to mention the sex of the holotype. 

description 
(1) anterior margin of clypeus not emarginate medially, obtusely biden- 
tate (Figs. 96, 135); (2) clypeal process reduced to small gibbosity; lateral 
clypeal carinae present; (4) cephalic process of female (and male?) a trans- 
verse, trituberculate carina (Fig. 135); paraocular areas of parietals flat, (5) 
wide; (6) postocular lobes of parietals depressed transversely, anterior por- 
tion of circumnotal ridge not emarginate; canthal area distinct, small; (7) 
cephalic brachium monopodal; (8) eyes small; (9) occipital ridge incom- 
plete; (10) labrum resembling Coprophanaeus-type (Figs. 140-141); (11) 
mentum weakly emarginate (Fig. 154); (12) median lobe of hypopharynx 
absent; labial palpi of one segment (Fig. 154) with membranous slit on 
dorsal surface (Fig. 155, arrow); (13) premental sclerites marginally sclero- 
tized; (15) pronotum of female (and male?) evenly convex, (16) evenly 
granulo-rugose and margined posteriorly by fringe of long setae behind a 
sharp, transverse carina (Fig. 178); posterior pronotal fossae absent; meso- 
metasternal suture angulate medially (as in Fig. 240) ; (17) front tarsi absent 
in female; (18) front tibiae quadridentate (Fig. 248); (19) middle and hind 
tarsi five-segmented, segments of characteristic shape (Figs. 266-267) ; dorsal 
surface of hind tibiae with dense, longitudinal fringe of hair (Fig. 266) ; 
(20) elytral interstriae punctate, (21) weakly convex, nearly flat; (22) striae 
simple; hind wing not notched (Fig. 296); last abdominal spiracle marginal; 
abdominal sterna 3-7 each produced as small acute tooth projecting from 
beneath elytral margin; abdominal sterna 3-7 with dense fringes of long 
hair, completely traversing sternum 7 (Fig. 301) while restricted laterally on 
sterna 3-6; length 10mm; color dull blue or green. 



Evolution of the Phanaeine Dung Beetles 855 

PART VI: EVOLUTION AND PHYLOGENY OF PHANAEINES 

Fossil phanaeines are unknown. Frenguelli (1938, 1939) attributed to 
Phanaeus ("o un genero afin") brood balls recovered from various Tertiary 
deposits (Oligocene through Pleistocene) in southern Argentina. 12 Pierce 
(1946), describing Palaeocopris from the La Brea tar pits (Los Angeles, 
California; Pleistocene), suggests a morphological resemblance to this genus 
with Phanaeus; I have not seen the type material of Palaeocopris. 

Lacking meaningful paleontological data, systematists concerned with 
evolutionary and phylogenetic questions are forced to base conclusions upon 
neontological data. The purpose of this discussion is to consider (1) the 
phylogenetic (cladistic) relationships among phanaeines indicated by an 
evolutionary interpretation of the comparative skeletal morphology and the 
biology of contemporary forms, and (2) the evolutionary implications of 
the ecology and distribution extant phanaeines. 

Some Preliminary Considerations 

Data for this discussion are essentially the same available for any 
approach to inferences about relationships, whether intended to be strictly 
phenetic or strictly cladistic. Since the data are the same, the conclusions of 
phenetic vs. phylogenetic inferences about relationships differ only as a result 
of the way they are treated. The evolutionary and phylogenetic conclusions 
arrived at here are based upon phenetic relationships indicated primarily by 
morphological characters and upon presumed cladistic relationships inferred 
by stressing (weighting) certain characters and by making judgments about 
evolutionary direction. 

Phylogenetic speculation requires some a priori notions about evolution- 
ary direction; that is, it requires decisions about which state of a character 
employed in assessing relationships is "primitive" (has undergone little or 
no evolution since the origin of a group; = plesiomorphous character of 
Hennig, 1965, 1966) and which is (are) "derived" (= "modified" = "special- 
ized"; apomorphous character of Hennig). Table 1 lists for phanaeines the 
presumed primitive and derived states for some of the characters used in this 
study which lend themselves to such judgment. Table 1 is, in effect, a list of 
evolutionary trends observable in phanaeines. It is to be noted, however, 
that while certain character states are designated as primitive for phanaeines 
(e.g., striking sexual dimorphism, metallic color), they may be specialized 
ones when considering Scarabaeinae as a whole. In these instances, the 
derived state in phanaeines can be a secondary (convergent) condition and 
phenetic equivalent of the primitive state for the entire subfamily. An im- 



12 Gonza!o Halffter (pers. comm.) has examined the brood balls used in Frenguelli's studies 
and concludes that they are essentially of the type presumed characteristic of phanaeines and 
isolated species of Duhotomius (see Halfiter and Matthews, 1966:130-139). 



856 



The University Science Bulletin 



Table 1. Primitive and derived features of phanaeines. 



Character* 


Primitive State 


Derived State (s) 


1. Shape of anterior margin of 


Not emarginate 


Emarginate medially 


clypeus 


medially 




2. Shape of clypeal process 


Transverse 


Spiniform, U-shaped or reduced 


A Lateral clypeal carinae 


Present 


Absent 


3. Shape of cephalic process of 


Corniform 


Lamellate, carinate 


male 






4. Shape of cephalic process of 


Carinate 


Corniform 


female 






5. Width of paraocular areas 


Narrow 


Wide 


6. Shape of postocular lobes of 


Depressed trans- 


Depressed obliquely — 


parietals — anterior portion of 


versely — 


emarginate 


circumnotal ridge 


not emarginate 




7. Shape of cephalic brachium 


Monopodal 


Bipodal 


8. Size of eyes 


Small 


Large 


9. Occipital ridge 


Incomplete 


Complete 


10. Type of labrum 


Phanaeiis-typc 


Copro p/uinaeus-iy pe 


1 1 . Shape of mentum 


Strongly emarginate 


Weakly emarginate 


12. Median lobe of hypopharynx 


Present 


Absent 


13. Premental sclerites 


Marginally sclerotized 


Completely sclerotized 


16. Pronotal sculpturing 


Smooth-punctate 


Roughened 


1 7. Front tarsi of female 


Present 


Absent 


18. Front tibiae 


P/ianaeus-type 


Copro p/ianaeas-type 
Dendropaem o«-type 


19. Middle and hind tarsi 


Five segmented 


With fewer than five segments 


22. Striae 


Simple 


Carinulate 


B Hind wing 


Not notched basally 


Notched basally 


C Position of last abdominal 


Marginal 


Intermediate, intrasternal 


spiracle 






D Color 


Bright (metallic) 


Sombre 


E Size 


Length under 30 mm 


Length over 30 mm 


F Secondary sexual dimorphism** 


Striking (sexes very 
different) 


Subtle (sexes similar) 


G Feeding habit 


Coprophagous 


Necrophagous 


H Time of activity 


Diurnal 


Nocturnal 



Numbering of characters corresponds to that in part III: reference should be made to this 
part for complete definition of morphological character states and their taxonomic dis- 
tribution. Those characters designated A-H are characters either discussed, but not num- 
bered, in part III or not discussed in that part. 
Based principally upon cephalic and pronotal shapes of well developed individuals. 



portant group of characters is not included in Table 1 : those which, because 
of their restricted taxonomic occurrence, are considered phanaeine special- 
izations (e.g., transverse clypeal carina of Diabroctis, the several unique 
features of Detidropaemon and its subtaxa, etc.). These specializations, which 
are obviously most useful in "keying" the taxa to which they are restricted, 
are taken into account in the following discussions. 

The direction of evolution is sometimes intuitively self-evident. Thus, a 
tarsal complement of fewer than five segments in phanaeines (and other 



Evolution of the Phanaeine Dung Beetles 857 

dung beetles) can (to me) be construed only as arising from the loss (or 
fusion) of one or more segments. In many instances, however, logical de- 
cisions depend largely upon a knowledge of variation in non-phanaeine 
taxa; for this reason it was necessary to examine other, non-phanaeine dung 
beetles (see Appendix for a list of those examined). When highly subjective 
judgments were called for in Table 1, the following points were considered: 
Since evolution as it is now understood tends to be a conservative process, 
a character state of widespread occurrence among otherwise quite diverse 
taxa is probably a primitive one. Accordingly, in character 7 the designation 
of a monopodal brachium as primitive was based principally upon the 
presence of monopodal brachia in all non-phanaeine scarabaeines examined 
as well as in the majority of phanaeines. Furthermore, the evolutionary 
status of the morphological character states regarded as correlates of be- 
havorial trends are most logically considered the same as the status of the 
behavioral phenomena in question. Thus, if necrophagy in phanaeines is a 
specialized behavioral pattern, character states regarded as morphological 
correlates of necrophagy (e.g., emarginate clypeus) are no doubt also derived. 
In making phylogenetic inferences I have stressed a fundamental hy- 
pothesis of "phylogenetic systematics" as most comprehensively approached 
by Hennig (1965, 1966): that, when practical, taxa should be recognized on 
the basis of shared derived character states, i.e., by synapomorphy or joint 
possession of apomorphous features (Hennig), since to do so increases the 
probability that taxa recognized are monophyletic. I do not, however, con- 
sider myself a phylogenetic systematist in the rigid sense of Hennig since I 
have chosen to temper the process of classifying and subsequent ranking of 
taxa with some measure of "taxonomic practicality." A "true" phylogenetic 
systematist (if such a person exists) might object to the phrase, "when prac- 
tical," above. Thus, the classification of phanaeines proposed above does not 
completely reflect my ideas about the phylogenetic relationships among 
phanaeines, which are discussed below. 

Phylogenetic Speculation 
Table 2 shows the taxonomic distribution of the character states of the 
26 characters listed in Table 1. For each taxon listed in Table 2 the per cent 
frequency of derived states (%D) is given at the right. 13 Using these per- 
centage figures it is possible to generalize about the relative overall "primi- 
tiveness" (or "derivedness") of the various taxa based on those 26 characters. 
Thus, taxa 3 and 4, the auricollis and imperator groups of Sulcophanaeus, 



13 In calculating %D, "O" entries were counted twice, once as primitive, once as derived; 
"X" entries were disregarded. Ordinarily %D = no. D's/26; but for, say, Oxysternon, sen. str., 

no. dots -j- no. O's 

%D = X 100%. 

26 + no. O's — no. X's 



858 



The University Science Bulletin 



V 4-j 

V u 

-o c 



G 




'rt 


CO 

N 


CO 

4— » 

G 


4-J 

-a 


CO 


o 


HJ 




3 


_C 




w 


+-< 


G 


4-J 
CO 


s 


~ 


c 






O 


O 






<-W 


-o 


bfl 




^-^ 


C 
O 


I 


a; 


r- 




4— » 


Ll 




rt 


ai 


1) 


oo 


CO 




-a 


4— ' 

cs 

4— > 


<u 


.•* 


CO 


— 






IT, 


a; 


T3 


CO 


~! 


IU 








«; 


II 




g; 


TD 


c 


J— 1 

O 


T3 


c 


G 


k. 


-o 


G* 


-o 


aj 


II 








o 


4—1 




on 




o 




£ 


_Q 


00 


Ul 


s 


Uh 


a- 


?^ 


<d 




CO 


4^ 


M-l 




u 


O 

;>> 

u 

G 


O 
G 

IT 


u 




«j 


n 


vo 


3 


rt 



H 9> 



4-» 


a: 
G 


u 


4-> 
CO 


<u 


G 


_G 

4-J 


II 


i- 
o 


o 


o 


G 


G 


^ — s, 


- 


o 


t-l 


G 


— 


IU 


-^ 


3 


(J 




_Q 


a! 


o 


!-. 


rt 


G 


— 


_G 


II 




U 


T3 




u 




X 








a 

o 

G 
O 


00 


G 

3 


X 




— 


« 


00 




H 


4-1 

« 

4-1 




rs 


00 




w 
m 


<u 
> 

• «— < 




< 


Ui 




H 








Evolution of the Phanaeine Dung Beetles 859 

emerge as most primitive (%D = 4), hence most obvious choices as the 
closest phenetic (and phyletic) approaches to "palaeophanaeines," while 
Dendropaemon (average %D = 58) appears to be the most highly evolved, 
hence most different from some hypothetical common ancestor. 

Table 2 is somewhat misleading since some groups (particularly taxa 
12-17) can actually be considered more highly evolved relative to the others 
than indicated by %D's. This is so since most of those character states unique 
to these taxa and, therefore, considered derived, are not enumerated per se 
in Tables 1 and 2. The incidence of such unique characters is especially high 
in Dendropaemon (e.g., body shape variation, lateral depression of base of 
pygidium, forked antero-median process of labral suspensorium, reduction 
in number of setae comprising median brush, etc.; see systematic section). 
The incorporation of these characters into Table 2 in the case of Dendro- 
paemon and other taxa would only intensify the phenetic gaps indicated 
there; the relative positions of all taxa on an "evolutionary scale" would not 
change. Incidence of "concealed synapomorphy" will be discussed more 
fully below. 

Of the 26 characters listed in Table 2, I have judged some intrinsically 
more useful than others for purposes of deducing phylogenetic relationships. 
Those presumed high in phyletic information content are shaded in Table 
2. Those characters not shaded (omitted from phylogenetic considerations) 
form two principal groups: (1) some characters (12, 17, 19, A) were ex- 
cluded because they are regressive, that is, their evolution involves loss or 
"reduction" (character 2, which is partially regressive, was not altogether 
ignored). I have accepted the hypothesis that regressive features (character 
states) are inherently more susceptible than progressive ones to parallel or 
convergent development (see, for example, Mayr, 1969: 222-223). Summary 
exclusion of regressive characters has probably resulted here in little loss of 
phyletic information since their inclusion would not significantly alter my 
phylogenetic inferences about phanaeines. (See discussion of parallelisms 
below.) (2) Characters D-H were excluded for discretionary reasons. Sepa- 
ration of largely discrete states of characters D, E and F is highly subjective; 
the designation of primitive states of characters G and H is based on unequal 
knowledge (e.g., the feeding habits of relatively very few Sitlcophanaeus are 
known, but these are coprophagous; on the other hand, relatively a large 
number of Phanaens are known to be coprophagous; as indicated by avail- 
able, but unequally representative data, I assume both groups to be basically 
coprophagous) . 

Taxonomic data on the phylogenetic (shaded) characters in Table 2 have 
been reorganized in Table 3 as a phyletic "affinity matrix" showing the 
frequency of shared derived character states (= frequency of synapomorphy) 
between all possible pairs of taxa. One can observe directly in Table 3 pat- 



860 



The University Science Bulletin 



Table 3. Affinity matrix (taxon X taxon) based on frequency of shared derived 
character states (synapomorphy) among pairs of phanaeine taxa. Taxa are 

numbered as in Table 2. 



Taxa 

1 

2 
3 
4 

5 
6 



9 
10 
11 

12 
13 

14 

15 
16 
17 



— Diabroctis 
, 

1 - Sulcophanaeus 

0- 



1 10- 



Oxysternon 



12 10 13- 



1 10 3 3 1- 



Phanaeus 



2 10 3 3 14- 



3 10 1 
10 1 
10 10 1 



6 — Coprophanaeus 

7 6- 



10 10 16 5 7 
10 10 15 5 6 
10 10 17 5 8 



7 - Dendropaemon 
10 7- 



10001012 2 3 4 4 4 - Megatharsis 

100010 12 2 3 4 4 4 4 - Tetramereia 

10 001012 2 3 4 4 4 4 4 - Homalotarstts 



2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 
Taxa 



terns of high and low frequency of synapomorphy between, among and 
within genera. All taxa (1-17) are presumed monophyletic, the constituent 
species (one or more) of which are themselves related by synapomorphous 
states described fully only in the classification proposed above. Open triangles 
along the diagonal in Table 3 embrace intrageneric comparisons for those 
genera with two or more subgeneric groupings. 

If, as is assumed, frequency of synapomorphy is a valid measure of in- 
ferred phyletic (cladistic) relationship, one should expect the relative fre- 



Evolution of the Phanaeine Dung Beetles 861 

quency of synapomorphy to complement the classification of the groups for 
which comparisons are made. That is, in the case of phanaeine taxa, one 
should expect (where applicable) the /«/rageneric frequency of synapomor- 
phy to be greater than intergeneric frequency, and that groups thus defined 
should be assignable to hierarchical groups largely corresponding with those 
of the classification proposed above. With some significant exceptions to be 
discussed below, this seems to be the general case. Inspection of Table 3 
reveals two rather distinct phyletic generic groupings and two enigmatic 
groupings together with some inferences about evolutionary rates. Phylo- 
genetic "conclusions" are presented below under two headings, generic 
affinities and parallelisms, and summarized in Figures 316 and 317. 

Generic Affinities 

Genus-group 1: Phanaeus-Oxysternon. A relatively close phyletic re- 
lationship between these genera is characterized by the presumed evolution- 
ary fixation of three derived features, the first two being unique to the genus 
group: a complete occipital ridge, completely sclerotized premental sclerites 
and bipedal cephalic brachia (seen otherwise only in some Sulcophanaeus). 
As indicated by %D in Table 2, Oxysternon (particularly Mioxysternon, 
%D = 28) is the more highly evolved of the genera. Tables 2 and 3, how- 
ever, do not account for synapomorphous states unique to these genera, most 
notable of which are the spinate metasternum and acute postero-median 
pronotal angle of Oxysternon and the head-prothorax interlocking mecha- 
nism of Phanaeus. 

The relatively low frequencies of synapomorphy (0-1) between this genus 
group and other taxa indicate phyletic isolation; that is, they support the 
hypothesis that this genus group can be regarded as a distinct evolutionary 
lineage. On the other hand, the relatively low incidence of synapomorphic 
states (3) characterizing the group (i.e., relative overall primitiveness) sug- 
gests a relatively slow evolutionary rate (compared to, say, genus group 2). 

A rather long phyletic history is indicated by the taxonomic pattern, i.e., 
the distinctness of taxonomic gaps. Thus, intermediate types minimizing 
gaps between genera, among subgenera and species-groups of Oxysternon 
and among the species complexes of the Phanaeus splendid ulus group are 
virtually non-existent. Conversely, the taxonomic difficulties arising because 
of the phenetic overlap among constituent subgroups of the P. vindex group 
suggests the hypothesis that this species group, (1) underwent relatively 
recent (and rapid ?) differentiation and/or (2) its differentiation, whether 
recent or remote, has been characterized by a comparatively low incidence 
of extinction. The geographic distribution of the vindex group (see below) 
supports the first alternative, that this species group underwent extensive 



862 The University Science Bulletin 

radiation in an initially restricted, though highly partitioned geographic 
theater (Mexico). 

Genus-group 2: Coprophanaeus-Dendropaemon. Relative to other 
phanaeines, these genera are highly evolved. Indeed, the extreme phenetic 
divergence of Dendropaemon makes it very difficult to conclude intuitively 
that there exists a close phyletic relationship between such different taxa as 
C. (Megaphanaens) lancijer (Figs. 181-182) and D. (Dendropaemon) nitidi- 
collis (Figs. 308-309) ! However, a close phyletic affinity between these 
genera is evidenced by a high frequency of synapomorphy (see Table 3) : 
emarginate clypeus and U-shaped clypeal process, Coprophanaeus-type la- 
brum, weakly emarginate mentum, and roughened pronotal sculpturing. 
Moreover, several other tendencies tend to reinforce the relationship (al- 
though they are not all considered in Table 3) : usual absence of front tarsi 
in females, similarly developed front tibiae, tendency for secondary sexual 
dimorphism to be less pronounced than in other phanaeines, specialized 
feeding habits (assumed of Dendropaemon), common occurrence of pro- 
notal ridging, as well as the conservation of a number of primitive aspects 
of head structure in otherwise quite highly evolved groups (monopodal 
brachia, persistence of distinct canthal areas, transverse depression of post- 
ocular lobes, incomplete occipital ridge, etc.). 

On the basis of the high degree of synapomorphy between Copro- 
phanaeus, sen. str., and Coprophanaeoides, I have concluded that the latter 
taxon is an annectant between the two genera. Moreover, it appears that 
Coprophanaeus, sen. str., is more closely related phyletically to Copro- 
phanaeoides than to either Megaphanaeus or Metallophanaeus. That Copro- 
phanaeoides and Dendropaemon, sen. str., are undeniably congeneric is 
supported by their sharing of a number of uniquely dendropaemine features 
(see above). Consequently, I must conclude that Coprophanaeus is a para- 
phyletic taxon (sensu Henning, 1966) and not a monophyletic one (Fig. 
317). 

A relative long history of the Coprophanaeus-Dendropaemon lineage is 
indicated by the phenetic isolation of infrageneric taxa as well as the occur- 
rence of "problematic" species. The origin of these taxonomic isolates is 
more logically interpreted as relict survival of more primitive types; these 
isolates include C. (Megaphanaeus) bellicosus, Metallophanaeus (which is 
itself likely not monophyletic; see systematic section) and Paradendro- 
paemon. The latter is most problematic since, because it lacks many of the 
salient features of Dendropaemon, it cannot be accommodated easily by a 
strict taxonomic delimitation of "Dendropaemon." Indeed, the superficial 
appearance of D. ganglbaueri is more like Tetramereia than any of its con- 
geners. Were it not for its possession of unmistakably dendropaemine 
mouthparts, I would be inclined to rank Paradendropaemon as a genus. 



Evolution of the Phanaeine Dung Beetles 



863 




phyletic 



distance ■ 



Fig. 316. Phylogram of the inferred phyletic relationships among species included in 
Sidcophanaeus. 



Moreover, the subgeneric rank of Paradendropaemon imposes the possibility 
that Dendropaemon , like Coprophanaeus, is paraphyletic. 

Taxa of uncertain affinities: The phyletic disposition of other phanaeine 
taxa is not as easily accomplished as that for the aforementioned genus- 
groups. Relatively speaking, Diabroctis and Sidcophanaeus are primitive 
groups most logically interpreted as comprising relict elements. (The dis- 
proportionately high %D of 28 of the S. faunas group is attributable to the 
differential possession of derived features [O's in Table 2] by the two con- 
stituent species-pairs, and not to apomorphous states shared by all four spe- 
cies; consequently, the %D of any given species in the group is less than 28). 
There are no derived states shared by all Sidcophanaeus; and, as measured 
by those characters considered in Tables 2 and 3, all species-groups are vir- 
tually isolated phenetically and, presumably, phyletically. Moreover, the 
intragroup variation is such that the phenetic gaps among most constituent 
species of the species groups of Sidcophanaeus are themselves quite distinct. 
I have concluded, therefore, that Sidcophanaeus is very likely a paraphyletic 



864 The University Science Bulletin 

(or polyphyletic) assemblage which brings together a number of phyletic 
isolates representing the remnants of numerous older, largely primitive 
phyletic elements (Fig. 316). This conclusion is supported by the known 
distribution of the genus (see below). What has been said of Sulcophanaens 
is also true of Diabroctis, which is evidently also a small relict group of 
phyletic rank equal to that of the species-groups of Sidcophanaeus. 

A somewhat similar situation exists with Tetramereia, Homalotarsits and 
Megatharsis, all of which are monobasic. For reasons discussed in the sys- 
tematic section, the two former genera should probably be merged. As a 
group, these species appear no more closely related to each other than either 
is to Dendropaemon, which they resemble only slightly more than Copro- 
phanaens (Table 3). In view of their extreme isolation within the phanaeines 
and of their often curious, unique specializations (e.g., one-segmented labial 
palpus of Megatharsis, unique shape of the tarsomeres of all three species, 
etc.), I have concluded that they represent highly divergent relicts of doubt- 
ful affinities with the Coprophanaeus-Dendropaemon lineage. A relictual 
status of Megatharsis and Tetramereia is also tentatively suggested by their 
distributions. 

Summary of phylogenetic speculation (Fig. 317): Phyletically, the 
phanaeine Scarabaeinae, which are presumed to comprise a monophyletic 
group, appear to consist of two distinct phyletic lineages, together comprising 
the bulk of known species, as well as a number of phyletically isolated, small 
groups representing relicts of several older elements. Such phyletic (phe- 
netic) fragmentation is at least compatible with the hypothesis that the 
phanaeines comprise an ancient group of origin and long-standing residence 
in South America. 

Parallelisms 

As can be seen in Table 2, for example, given apomorphous character 
states are not necessarily restricted to presumably monophyletic taxa. In these 
instances one must ordinarily assume that the derived state arose inde- 
pendently in two or more phyletic lines. The origin of the similarity of two 
(or more) taxa resulting from joint possession of the same, independently 
acquired derived state (s) can be interpreted as the result of either parallel 
evolution or convergent evolution. Differences between these concepts are, 
to me, more in degree (taxonomic frame of reference) rather than in sub- 
stance since overemphasis of either kind of similarity will result in the recog- 
nition of polyphyletic groups. For present purposes, the presumed inde- 
pendent expressions of a given apomorph among phanaeine taxa are here 
cited as parallelisms while those few cited instances of synapomorphy be- 
tween phanaeines and non-phanaeines are considered convergences. 

Since I am not familiar with a large enough taxonomic variety of 



Evolution of the Phanaeine Dung Beetles 



865 



U 

c 




o 

c 

TO 



— 



n 
u 



•s 



i! 

c 



oo 
.C 

°C 
u 

£ 

3 



X 

a 

o 

a 

'v 
a 
c 
a 

c 

-j 

'G 



3 

•Si 

bo • 
C 2 

o g 

.£* a 
IS « 

c 

Q 



-a 
.s c 



■2 .S 

u 
a. j« 

.9 -a 

u 

"3 XI 



O 3 
_ C 

2 S 



CO 



S'-S 



ro 



o -a 
- «j 
b -a 

CO 



866 The University Science Bulletin 

Scarabaeinae to compile a comprehensive list, I have limited a discussion of 
convergent synapomorphy to mention of the following few instances: 

a. transverse clypeal carina: Diabroctis and Onthophagits. 

b. apical hind tibial carina: various phanaeines (see below) and "Co- 
prina" (Copris and related genera). 

c. lack of tarsal claws: all phanaeines, Eucranini and Onitini. 

d. lack of front tarsi in males: all phanaeines, some Onitini, some Can- 
thonia. 

e. reduction in tarsomere number: Dendropaemon, Heteroditopus (= 
Pinacotarsus) . 

f. elongate, flattened body: some Dendropaemon, Onthochans, Stipto- 
podius. 

g. necrophagous habit: Coprophanaeus and other species (see below), 
Deltochilum, some Dichotomius, some Cant/ion. 

h. P ha nae its-type antennal club: all phanaeines, Sceliages. 

Instances of presumed parallelisms among phanaeines are quite numer- 
ous. While the following list is not exhaustive, it will serve to illustrate the 
variety of characters affected and taxa involved. The name of each derived 
state is followed by those phanaeine taxa within each of which the apomorph 
is assumed to have arisen. 

a. presence of transverse hind tibial carina: Oxystemon jestivum group; 
Diabroctis; Sidcophanaeits carnijex-jaunus pair; S. imperator-steinheili 

pair. 

b. large size (length greater than 30mm): Diabroctis; Megaphanaeus; 
Sulcophanaeus f annus group. 

c. reduction of tarsomere number: Dendropaemon; Tetramereia; (? 
Paradendropaem on ) . 

d. bipodal cephalic brachium: Phanaeus-Oxysternon genus-group; Sidco- 
phanaeits imperator; S. laeander-cohtmbi pair. 

e. corniform female cephalic process: Megaphanaeus; Sidcophanaeits 
faunits group. 

f. intrasternal position of last abdominal spiracle: Megaphanaeus; Sulco- 
phanaeus faunus-carnifex pair; Mioxysternon. 

g. loss of front tarsi in female: Metallophanaeus; Coprophanaeus, sen. 
str.; Tetramereia-Homalotarsus; Sidcophanaeits carnifex; S. steinheili. 

h. flat elytral interstriae: Sidcophanaeits auricollis group; Megaphanaeus; 

Phanaeus damocles-quadridens; P. vindex-difformis. 
i. rugose elytral interstriae: Megaphanaeus; Phanaeus vindex complex, 
j. carinulate striae: Megaphanaeus; Dendropaemon; Phanaeus bis pin us; 

P. \irbyi-palaeno pair, 
k. notching of hind wing: Coprophanaeus, sen. str. -Dendropaemon; 

Phanaeus bispinus. 



Evolution of the Phanaeine Dung Beetles 867 

1. roughened pronotum: Coprophaneus-Dendropaemon genus group; 

Phanaeus vtndex group; Sulcophanaeus laeander-columbi pair; Tetra- 

m ereia-Ho m alo tar sits ( -Megath arsis ) . 
m. spiniform clypeal process: Oxystemon festivum group; Phanaeus 

kjrbyi-palaeno pair, 
n. necrophagous habit: Coprophanaeus; isolated species of Phanaeus 

(e.g., bispinus, meleagris); isolated species of Sulcophanaeus (e.g., 

faun us). 

Evolutionary Implications of Phanaeine Ecology and Distribution 

Halftter and Matthews (1966) discuss quite fully various aspects of the 
ecology and behavior of Scarabaeinae. Rather than redevelop their discussion 
as it applies to phanaeines, comment here will be limited to the following 
summary statements: 

1. The principal food sources of phanaeines are animal feces (particu- 
larly the dung of large herbivores such as cattle) and carrion. 

2. Necrophagy as such is likely a behavioral adaptation for forest ex- 
istence and a derived behavioral pattern in Scarabaeinae (see Halfifter and 
Matthews for a detailed argument). Dung beetles in South American 
forests are very often necrophagous; conversely, few necrophagous species 
inhabit grasslands. This behavorial pattern usually cuts across established 
taxonomic lines, such that very few taxa can be delimited by necrophagous 
feeding behavior as well as by morphology. Coprophanaeus is one of these 
few. Ecological experiments designed to test the relationship between nec- 
rophagy and forest habitat have not been done. Among other things, these 
might investigate (a) the apparent scarcity of carrion in open grassland 
compared to a forest, (b) rates of desiccation of carrion in a forest vs. grass- 
land, and (c) the frequency and severity of competition between necro- 
phagous animals in a forest vs. in a grassland (e.g., between beetles and 
vultures in grasslands). 

3. Within the phanaeines the taxonomic distribution of principally 
diurnal-coprophagous-grassland species and nocturnal-necrophagous-forest 
species is largely generic. Thus, most Phanaeus are coprophagous-grassland 
forms while Coprophanaeus consists entirely (as now known) of necro- 
phagous-tropical forest species. These two genera include about 60% of all 
phanaeines. Diabroctis, Sulcophanaeus and Oxystemon are poorly known 
ecologically, but all appear to be largely coprophagous groups with both 
grassland and forest species. Nothing is known about the biologies of 
Dendropaemon, Tetramereia, Homalotarsus and Megatharsis. 

Phanaeines are restricted to the New World, occupying South America 
(except Chile and southern Argentina), Central America, Mexico and east- 
ern and extreme southwestern portions of the United States (Maps 1-6). The 



868 The University Science Bulletin 

Table 4. Summary of the distribution and relative abundance of phanaeines.* 









Region 






Taxon 


1 


2 


3 


4 


5 


Diabroctis 


(+) 


— 


— 


— 


— 


Sulcophanaeus 


( + + + + ) 


( + ) 


+ ( + ) 


+ 


— 


Oxysternon 


( + + + + + ) 


— 


— 


— 


— 


Phanaeus 












spletididulus complex 


( + + ) 


(++) 


(1 s P ) 


(2 sps) 


— 


vindex complex 


— 


— 


(+) + 


( + + + + ) + + 


+<+) 


Coprophanaens 












Megaphanaeus 


(+) 


— 


— 


— 


— 


Metallophanaeus 


(+) 


— 


— 


— 


— 


Coprophanacus, s.s. 


( + + + + + ) + 


1 sp 


+ 


+ 


— 


Dendropaemon 


(+++++) 


— 


— 


— 


— 


Megatharsis 


(1 sp) 


— 


— 


— 


— 


Tetramereia 


(lsp) 


— 


— 


— 


— 


Homalotarsus 


(1 sp) 


— 


— 


— 


— 



* See text for explanation. 

only insular species is Sulcophanaeus carnijex, which is endemic to Jamaica. 
Table 4 summarizes the distribution according to the following numbered 
regions : 

1. Tropical South America (except northwestern sector and areas west of 
the Andes) 

2. Temperate South America 

3. Northwestern South America (including areas to the Pacific side of 
the Andes) and Panama 

4. Central America (except Panama) and tropical Mexico 

5. Temperate Mexico and the United States. 

In Table 4 presence in one or more regions is denoted by one or more + 's, 
or by the number of species if very few; — denotes absent. The number of 
+'s indicates relative abundance in a given region; parentheses enclose that 
portion endemic to the region. 

The tropical South American fauna includes representatives of all 
phanaeine genera, eight of which are endemic or largely so. Such a high 
degree of generic endemism strongly suggests South America as the place of 
origin of the group. Differentiation of all genera (as presumably could also 
be the case for other South American scarabaeine groups) was probably 
before or during Tertiary isolation of South America. If during Tertiary 
isolation, the greatest development of phanaeines could have coincided with 
that of South American herbivores (e.g., the ungulates). During periods of 



Evolution of the Phanaeine Dung Beetles 869 

isolation inter-American water gaps presumably made difficult northward 
dispersal of both dung beetles and large land animals with which they 
were associated. If the initial development of scarabaeine groups in all geo- 
graphic areas were closely tied to the presence of a large food supply pro- 
vided primarily by the dung of large animals, the extinction from South 
America of grassland herbivores must have greatly affected the evolution of 
the scarabaeine fauna. Necrophagous, forest dwelling groups, such as Copro- 
phanaeus, could have resulted from movement of grassland groups into 
forest habitats in response to extinction of herbivores. The coprophagous 
groups which survived were probably largely restricted to small, geograph- 
ically isolated refugia where they differentiated as small, phenetically quite 
dissimilar groups. 

The disjunct patterns of distribution within the faunas and imperator 
groups of Sulcophanaeus are indicative of relict groups and, hence, com- 
patible with the hypothesis presented above that the genus comprises the 
relicts of several older phyletic elements. The following disjunct patterns 
between closely related species are not evident in Map 1: 

a. Sulcophanaeus f annus, Amazon Basin — S. carnife.x, Jamaica. 

b. S. laeander, Venezuela (savannas?) — S. columbi, southern Amazon 
Basin. 

c. S. imperator, Andean highlands of northern Argentina and of Bolivia — 

S. steinheili, high mountain valleys of Colombia. 
The distribution of the auricollis group (Map 1) is evidently continuous. 
This appears to be the only sulcophanaeine group that has undergone even 
limited northward dispersal and differentiation. The close resemblance of 
the Andean species auricollis and the Mexican chryseicollis and the occur- 
rence of two species (noctis and velutinus) in both South and Central 
America suggest that the movement was relatively recent, perhaps accom- 
panying the equally limited northward dispersal of Coprophanaeus, sen. str. 
(Map 5) and the Phanaeus splendidulus group (Map 2) (see below). 

While the relict status of Megatharsis and Tetramereia (and Homalo- 
tarsus) is supported by the extreme phenetic and geographic isolation of 
these species (Map 6), both are very poorly represented in collections and 
may later prove to be more widely distributed (in the case of Megatharsis) 
or more evenly distributed (Tetramereia). Likewise, the distribution of 
Dendropaemon is poorly known and, therefore, not amenable to serious 
geographic analysis. 

Within other genera distributional patterns appear more regular, i.e., 
with closely related species occupying largely adjacent regions. There is a 
tendency, however, for phenetic isolation to be coincident with geographic 
isolation (e.g., in the cases of Coprophanaeus pertyi and pessoai, C. belli- 
cosus, and the species of Mioxysternon). The northward dispersal and 



870 The University Science Bulletin 

accompanying linear differentiation (speciation) of the S. auricollis group 
is paralleled in Coprophanaeus, sen. str., by the dardanus group (C. telamon 
extending from South America into Central America and southern Mexico; 
C. pluto endemic to Mexico) and in the Phanaeus splendid iilus group by the 
endymion complex (P. blanchardi in northwest South America, pyrois in 
Central America and endymion in Central America and southern Mexico). 

The Phanaeus find ex group is the only phanaeine taxon centered geo- 
graphically outside South America. This large species-group is centered in 
Central America and adjacent southern Mexico, where it almost certainly 
originated and underwent extensive, probably rapid radiation (as indicated 
by the narrow, taxonomically inconclusive phenetic gaps between clusters of 
closely related species). From Mexico the group dispersed both northward 
and southward, differentiating further in both directions, although most 
clusters remained in or near the locus of radiation. Northward movement 
of Mexican species undoubtedly gave rise to a temperate fauna (the vmdex 
complex) largely endemic to the United States. Although the possibility 
needs to be examined more closely, the distributions of U.S. species indicate 
that their history includes disruptions by Pleistocene glaciation (suggested by 
my somewhat premature suspicion that P. floridanus and isolated popula- 
tions of P. vindex in Arizona are products of differentiation in refugia from 
glaciation). The hermes cluster, conversely, likely arose from South Amer- 
ican immigrants moving southward from Mexico. Apparently the only 
group to disperse successfully in both directions was the mexicanus cluster. 
The combined range of the constituent species of this cluster extends from 
Peru (lunaris) through Central America (ivagneri), southern Mexico (mexi- 
canus), and along the western coast of Mexico into Arizona (amythaon). 
The ecological-geographical distribution of Phanaeus in Mexico is quite 
complex and in need of continued analysis. 

In trying to assess the significance of geographical patterns of phanaeine 
distribution, two points come to mind. First, field work in areas like the 
Amazon Basin is restricted logistically to efficient avenues of penetration, 
i.e., mostly large rivers. Thus, one can gather data yielding sound ideas 
about beetles collected along and near rivers but no ideas about what occurs 
in areas between rivers, which in the Amazon Basin are immense pieces of 
real estate. Consequently, inferences about distributions of largely Ama- 
zonian groups are subject to the hypothesis that, because the area appears 
monotonously "homogenous" from an airplane, the unworkable inter-river 
areas may be assumed to be faunistically similar to areas along and near 
encompassing rivers. This is a logical, and probably largely accurate ex- 
pectation. But just how "homogenous" are the 2.5 billion or so acres called 
the Amazon Basin? Second, an important consideration is the possible ef- 
fects of human activity on the recent movements and abundance of dung 



Evolution of the Phanaeine Dung Beetles 871 

beetles. The livestock industry in all parts of the world has undoubtedly 
resulted in the geographic expansion and increased abundance of a great 
many coprophagous forms. Seemingly, the faunas least affected by such 
activity would have been those supported prior to the introduction of live- 
stock by indigenous herbivores, such as the bison in the North American 
Great Plains and the rich herbivore fauna of African savannas. The positive 
effects of the introduction of livestock has no doubt been great in Latin 
America, which in recent, pre-European times supported a very depauperate 
indigenous herbivore fauna. Many coprophagous species of phanaeines and 
other dung beetles now considered very common were perhaps quite scarce 
in those times. A small scale example is the case of Sulcophanaeus carnijex, 
which today can be collected in incredible numbers in cow dung on Jamaica 
(Matthews, 1966) ; Jamaica has presumably never supported large herbivores 
other than cattle, the only known mammalian fossils being bats, monkeys 
and small rodents. On the other hand, the same livestock producing activ- 
ities, which always involve the cutting of forests for pasturage (or timber, 
crops, etc.), has undoubtedly had some negative affect on the distribution 
and abundance of stenotopic forest species, whose habitat is shrinking at a 
tremendous rate in many parts of Latin America. 

LITERATURE CITED 

Arrow, G. J. 1951. Horned Beetles. Junk, Lochem. 154 pp. 

Balthasar, V. 1942. Uber die Nervatur des Hinterfliigels tier Lamellicornien. Ceske 

Sploecnosti Nauk Trida Matematicko-Prirodovedecka Vestni'k 1942, Part XI, 1-44. 
Barattini, L. P., and A. Saenz. 1961. Contribucion al conocimiento de las especies del 

genero Gromphas Brulle. l ero Congr. Sudamericano Zool. 3:21-29. 
. 1964. Contribution a l'etude des especes du genre Gromphas Brulle. Rev. Franc. 

d'Entom. 31:173-181. 
Blackwelder, R. E. 1944. Checklist of the coleopterous insects of Mexico, Central America, 

West Indies and South America, part 2. United States National Museum, Bulletin 185, 

pp. 189-341. 
Blut, H. 1939. Beitrag zur Verbreitung and Svstematik der Gattung Dendropaemon. 

Archivs fur Naturgeschichte (N.S.) 8:263-300. ' 
Brulle, M. 1837. Observations sur l'absence des tarses dans quelques insectes. Ann. Sci. 

Natur., Zool., ser. 2, 8:246-249. 
Butt, F. H. 1944. External morphology of Amphimallon majalis (Razoumovski) (Coleoptera, 

the European chafer). Cornell Univ. Agric. Exp. Sta. Memoir No. 266. 18 pp. 
Castlenau, Gomte de. {■=. F. de Laporte). 1831. (Description of Enicotarsus vrndipennis) . 

Mag. Zool. 1 (2) :plate 35. (Reference from Blackwelder, 1944.) 

. 1840. Historie Naturelle des Insectes, Coleopteres, vol. 2. Paris. 564 pp. 

Crowson, R. A. 1938. The met-endosternite in Coleoptera; a comparative study. Trans. Rov. 

Ent. Soc. London 87:397-416. 
. 1944. Further studies on the met-endosternite in Coleoptera. Trans. Roy. Ent. Soc. 

London 94:273-310. 
. 1955. The Natural Classification of the Families of Coleoptera. N. Lloyd & Co., 

London. 187 pp. 
Dejean, P. 1837. Catalogue des Coleopteres de la Collection de M. le comte Dejean. 3rd ed. 

(rev. corr. et augm.). M.-M. Pere et Fils, Paris, 384 pp. 
. 1838. (. . . observations . . . sur l'absence ou la presence des tarses dans plusieurs 

insectes de la famille Lamellicornes . . .) (from the minutes of a society meeting.) 

Ann. Soc. Ent. France 1838, pp. viii-ix. 
Dorsey, C. K. 1943. The musculature of the labrum, labium and pharyngeal region of adult 

and immature Coleoptera. Smiths. Misc. Coll. 103(7): 1-42. 



872 The University Science Bulletin 

DuPorte, E. M. 1960. Evolution of cranial structure in adult Coleoptera. Canad. J. Zool. 

38:655-675. 
Edmonds, W. D. 1967. The immature stages of Phanaeus {Coprophanaeus) jasius Oliver and 

Phanaeus (Metallophanaeus) saphirinus Sturm (Coleoptera: Scarabaeidae). Coleopt. 

Bull.